The frame transfer CCD with high sensitivity, high resolution and wide spectral response is the main image sensor of the satellite-borne polarization camera. The image fringe blurring phenomenon caused by the inherent structure of the frame transfer area array CCD is called smearing, which will reduce the imaging quality and spectral measurement and then affect the accuracy of remote sensing image product parameter inversion. Therefore, research on the evaluation standard and correction method of the tailing degree is of great significance to realize the high-precision measurement of satellite-borne polarization camera. Based on the working principle of the directional polarization camera, which assembles the frame transfer CCD, this paper analyzes the mechanism of smearing and introduces the dark line correction model by using the dark line of frame transfer CCD. According to the characteristics of the dark line correction model, this paper analyzes the light leakage problem of some dark line pixels adjacent to the photosensitive area due to the point diffusion effect, and it is concluded that different dark lines have an effect on the smearing correction effect. To avoid the problem of inaccurate correction of smearing caused by light leakage dark lines, this paper proposes a method to compare the degree of image smearing to select the appropriate number of dark lines. That is, the degree of smearing of the image corrected by different numbers of dark lines is evaluated by appropriate criteria. After comparison, the image with the lowest smearing degree after correction is selected as the image with the best correction effect, and the corresponding image is the most appropriate number of dark lines. To find suitable evaluation criteria, the smearing simulation model of the integrating sphere imaged by the directional polarization camera is established by using MATLAB. Combined with the smearing simulation model, the gray standard deviation and average gradient of the smearing area are used as the evaluation criteria of the image smearing degree. The smearing stripes with different brightnesses of the integrating sphere are obtained by using the simulation model, and then the gray standard deviation and average gradient in the smearing area are calculated. Then, the experiment of imaging different luminance integrating spheres by directional polarization camera is carried out, the same brightness as the simulation integrating sphere is set, and the experimental images under different luminances of integrating sphere are obtained. The gray standard deviation and average gradient of the smearing area in the experimental image are calculated to verify the simulation data. The results show that the simulation data are in good agreement with the experimental data, which verifies the effectiveness of the evaluation criteria. Finally, to solve the problem of inaccurate smearing correction caused by light leakage dark lines, a correction smearing optimization algorithm is proposed according to the evaluation criteria of the gray standard deviation and average gradient, which is verified by integrating sphere imaging at the edge of the field of view of the directional polarization camera. The algorithm can calculate and compare the gray standard deviation and average gradient of the smearing image corrected by different dark lines in each band. The experimental results show that the correction smearing optimization algorithm can adaptively determine the most appropriate number of dark lines needed to correct smearing in the current experimental environment. Using this algorithm, the gray standard deviation percentage of the image smearing after the correction of each channel of the directional polarization camera is reduced by 50.3%~89.9%, and the average gradient percentage is reduced by 65.2%~74.0%. The research results provide a reference for the detection and correction of image smears by satellite-borne polarizing cameras.
Photodetection is an important means of modern detection and intelligent sensing technology. Classical photodetectors are based on the interband transition of electrons in the semiconductor materials, and their maximum response wavelength depends on the optical band gap of the semiconductor materials. Only photons with energy higher than the band gap (wavelength less than the maximum response wavelength) can be detected. Infrared photoelectric detectors are widely used in military reconnaissance, aerospace remote sensing, astronomical observation, industrial detection, optical fiber communication, infrared imaging and other fields. Common infrared detectors are made of semiconductor materials with extremely narrow band gap, which are facing many problems such as complex processes, high costs and low operation temperature. Hot electrons photodetectors based on the internal photoelectric effect have attracted tremendous attention in the past decade due to its advantages such as overcoming the response limitation of the semiconductor band gap, fast response speed, working at room temperature and light polarization sensitivity. The main problem of hot electrons photodetectors is low device responsivity, especially in the near infrared regions. In recent years, numerous studies at home and abroad have demonstrated that optical means such as grating plasmons, Tamm plasmons and micro-cavity effect can effectively enhance the optical absorption of metal films in the hot electrons photodetectors, and further improve the device responsivity. The planar metal thin films / Distributed Bragg Reflector (DBR) Tamm plasmons have many advantages such as simple structure, low manufacturing cost, high absorption efficiency. By changing the DBR structure parameters and metal thin films thickness, the resonance wavelength and maximum response wavelength can be adjusted. This structure provides an effective way to improve the absorption of hot electrons photodetectors, but the corresponding devices normally exhibits narrowband absorption and responsivity. The development of hot electrons photodetectors with broadband response characteristics is conducive to the expansion of its applications in optical fiber communication, photocatalysis, solar cells, solar water splitting and other fields. In view of broadband hot electrons photodetector, there are still some issues to be solved, such as narrow response spectra, low absorption efficiency, and low responsivity. TiN has high dielectric constants and excellent plasmons characteristics in the near infrared ranges. Moreover, the mean free path of hot electrons in the TiN thin films is about 50 nm, which is larger than the values of Au and Ag. Therefore, TiN is considered as potential candidate which can be used to achieve high performance broadband hot electrons photodetectors. In this study, we propose a multi-layer device architecture based on the TiN/TiO2 Schottky barrier and the Tamm plasmons formed by the TiN/ DBR structure to achieve high-performance broadband near infrared hot electrons photodetectors. This architecture has the following characteristics: 1) The TiN/DBR Tamm plasmons with manipulating the DBR structural parameters are able to enhance broadband absorption of the TiN thin films and broaden the device response spectra; 2) The Schottky barrier of TiN/TiO2 is only 0.37 eV, which will be beneficial to the transporting of low energy hot electrons; 3) The MgF anti-reflectance layer is used to reduce optical loss. Based on the optical transfer matrix and hot electrons emission theory, we firstly simulated the reflection and absorption spectra of the hot electrons photodetectors. Furthermore, we calculated the corresponding device responsivity. The simulation results show that with high refractive indices ratio dielectric layers, the DBR form by Ge/SiO2 can effectively expand the absorption spectra of the TiN thin films and the device response spectra. The absorption of TiN thin films and device responsivity also can be largely enhanced. By adjusting the DBR structural parameters, such as dielectric layers, DBR period and the DBR center wavelength, the absorption spectra of TiN thin films and the responsivity spectra of the hot electrons devices can be regulated. We also discuss the influences of the thickness of TiN thin films and MgF anti-reflectance layer. The maximum absorption of the TiN thin films is 99.8%, and the maximum responsivity value of the optimized device is 29.2 mA/W. The full width at half maximum of the response spectra is about 900 nm. The simulation results indicate that as the incident angle increases from 0 to 30°, the Tamm state can still be excited, but the peak absorption of the TiN thin films and the device responsivity are decreased. The peak absorption and response wavelength under the TE polarization and TM polarization are both blue shifted with the increasing of incident angles. This multi-layer device architecture provides a new route to realize high-performance broadband near infrared hot electrons devices. Meanwhile, it is beneficial to expand the applications of hot electrons photodetectors.
Just as humans use two eyes to obtain 2D and 3D visual information of the scene to move and work freely in the objective world, robots also need RGB 2D color images and 3D point cloud information of the scene and target when carrying out target recognition and fine work. At present, Microsoft Kinect, Intel R200, R461, and other RGB-D cameras on the market can meet the needs of robot applications in the air. Underwater robots also need RGB-D vision sensors to complete underwater target recognition and fine operation tasks, but so far there are no reports of commercial products of underwater RGB-D cameras. At present, the 3D measurement technology based on binocular stereo vision technology in the air has been relatively mature. However, the particularity of the underwater environment brings difficulties to the development of underwater RGB-D cameras. Firstly, due to the absorption and scattering of water medium and suspended particles in water, the captured underwater image has more serious problems than the image in the air, such as color distortion, low contrast, fuzzy details, which affects the matching accuracy of corresponding points in the stereo matching process and the authenticity of RGB image color. Therefore, for underwater RGB-D camera, underwater image enhancement processing is particularly important. The image enhancement technology studied by many scholars provides high-quality image guarantee for underwater stereo vision. In addition, when the industrial camera is applied to the underwater scene, the camera must be encapsulated in a waterproof shell. The propagation of light needs to pass through three media of "water-glass-air" in turn, and finally image in the camera image plane. However, due to their different refractive indices, light is refracted at the interface of different media, so that the object-image relationship of underwater camera imaging does not meet the imaging principle in air. Therefore, the traditional camera model is invalid. Researchers around the world have proposed that building an underwater refraction model is the fundamental solution to the problem of underwater 3D reconstruction. Although the research on underwater RGB-D technology is rare, the above research has paved the way for underwater RGB-D camera. In order to meet the needs of underwater target recognition and fine operation of an underwater robot, an RGB-D underwater camera engineering prototype based on binocular stereo vision integrating 3D point cloud and 2D RGB image are developed, and the related technologies are studied. It mainly includes: Firstly, considering the underwater multi-layer refraction effect, the underwater camera imaging and binocular stereo imaging models are established to convert the underwater image into the image in the air to eliminate the influence of refraction on the reconstruction accuracy. Then the corresponding points are matched to obtain the 3D point cloud; Secondly, according to the principle of underwater image imaging and Retinex theory, a color correction algorithm based on the attenuation correction coefficient of the underwater image is proposed to correct the color of the underwater image, and then the underwater uniform light component estimation method is improved based on the dark primary color prior theory to enhance the definition of the underwater image. Several common image enhancement algorithms are selected as a comparison, and the information entropy, the number of SIFT feature points, and canny feature points are used as evaluation indexes to evaluate the effect of different algorithms; Finally, the RGB values of the three channels of the matching points are extracted, and the standard RGB values are calculated by using the RGB information of the left and right matching points. The alignment superposition model is established, and the 3D point cloud and 2D color data are superimposed and fused to obtain underwater RGB-D image data. The experimental results show that the proposed underwater image enhancement algorithm can not only correct the color distortion of underwater image, but also improve the effect of underwater 3D reconstruction, which lays a foundation for obtaining higher quality RGB-D data; the established point cloud and RGB color registration model can register the color corrected color data with 3D point cloud data to generate high-quality underwater RGB-D data; the underwater RGB-D engineering prototype has good 3D measurement accuracy, and the system reconstruction error at 3M is 2.6 mm, which has a good actual underwater 3D reconstruction effect.
Laser has been widely used in laser medical treatment, radar, laser guidance, welding, and other fields. Many applications require specific modes, and the laser beam mode generated by the laser is usually Gaussian. For these particular cases, beam shaping is indispensable. The Diffractive Optical Element (DOE) is small in size, light in weight, easy to control the wavefront, and has good beam shaping performance. There are many ways to design DOE, such as genetic algorithms, simulated annealing algorithms, geometric algorithms, and iterative Fourier transforms. For optical design, a single algorithm is usually not enough. For example, genetic algorithms have strong global optimization capabilities and weak local optimization capabilities. The global optimization ability of the simulated annealing algorithm depends on the annealing temperature. If the temperature drops too fast, it is easy to skip the minimum value, and if the temperature drops too slow, the optimization time is too long. The iterative Fourier transform algorithm has the advantages of being fast and efficient and is an ideal choice for DOE design, but it is easily affected by the initial stage, leading to poor final results. A single algorithm cannot obtain satisfactory design results due to inherent defects. Some improvements on single algorithms or hybrid algorithms, such as the genetic simulation hybrid algorithm, fuzzy control iterative algorithm, and double-constrained GS improved algorithm, can compensate for the above shortcomings.To improve the efficiency and light quality of the optical diffractive element, an improved GS algorithm is proposed, in which the initial phase is introduced, and an iterative optimization algorithm is employed to obtain the best parameters. The iteration and neighborhood optimization algorithms are adopted to obtain the best DOE phase. The proposed design method can effectively improve the beam quality and diffraction efficiency. The DOE phase-by-phase unwrapping algorithm can generate a continuous surface suitable for diffractive devices manufactured by moving mask technology and reduce the machining errors introduced by quantization. In the simulation results, the diffraction efficiency is 0.999, and the root mean square error is 0.018 7. The experimental speckle contrast is approximately 0.011 8. In addition, the influence of four factors on the quality of the output spot is discussed: 1) The proposed phase parameter has a significant influence on the quality of the output spot, and there is an optimal value that can meet the requirements of high diffraction efficiency and low root mean square error. The simulation results of the light spots with the phase parameter d equal to 40, 62, and 90 are given here. When phase parameter is 40, there is a large bulge in the center of the output point. When phase parameter is 90, the output spot has noticeable depressions; when phase parameter is 62, the quality of the spot is the best, with neither protrusions nor depressions. 2) The waist radius of the incident Gaussian beam is also an important factor affecting the output spot. The simulation results for waist radii of 650 μm, 960 μm, and 1 550 μm are given. When waist adius is 650 μm, the uniformity of the output light spot decreases, and some fringes appear. When waist adius is 1 550 μm, the output light spot has prominent fringes, and the energy of the light spot drops significantly. When waist adius is 960 μm, the output spot uniformity is good, and most energy is concentrated in the target area. There is an optimal positional relationship between the incident Gaussian beam and the DOE. 3) The center of the incident Gaussian beam should be aligned with the center of the DOE. If it deviates from the optimal position relationship, the output light spot will be greatly affected, and the quality of the light spot will be reduced. This paper presents the simulation results of two horizontal directions with center deviations of -1.024 mm, 0 mm, and 1.024 mm. The right direction is defined as the positive direction. It can be seen from the figure that the deviation of the two centers causes streaks in the output light spot, and the uniformity is reduced to a certain extent. 4) Two random disturbances of 0.10π and 0.01π were added to the final DOE phase distribution and compared with the undisturbed results. It was found that the diffraction spot did not change significantly under the random disturbance of 0.01π, which was almost the same as the original spot. Even if a random disturbance of 0.1π is added, the shape of the diffracted spot remains good; only an inevitable decrease in the internal uniformity and speckles appear in the target signal area.
The wavelength of terahertz wave is between those of infrared wave and microwave, with the characteristics of penetration, strong water absorption, and low photon energy, etc. Terahertz imaging has been are widely used in biomedical imaging, anti-terrorism security and non-destructive testing and other fields. The good penetration of terahertz wave is beneficial to detect the internal structure of objects. Phase-contrast imaging can better reflect the internal structure of objects, compared with simple terahertz intensity imaging. Continuous terahertz wave phase-contrast imaging can be divided into point-by-point scanning acquisition method with a single-point detector and full-field recording method with a surface array detector. The former has a low data acquisition rate and cannot image large samples quickly; the latter has a fast data acquisition rate, but also places higher demands on the detector's pixel size and dynamic range. Continuous terahertz wave digital holographic imaging records the hologram by a faceted detector, and then obtains the complex amplitude information of the sample by a reconstruction algorithm. According to different optical path structures, continuous terahertz wave digital holography can be divided into in-line digital holography and off-axis digital holography. In-line digital holography can make full use of the spatial bandwidth product of the detector, but requires a sufficiently high energy ratio of direct and diffracted light after passing through the sample, as well as the problem of twin image interference. Off-axis digital holography introduces a certain angle between the object light and the reference light, which solves the twin image problem and makes the reconstruction process simple.Off-axis image-plane digital holography uses a microscope objective to directly collect the imaging information of the object. Compared with off-axis Fresnel digital holography, this imaging method has fast computation and high resolution, and has been widely applied at visible wavelengths. Besides, this imaging method has been combined with some other high-resolution imaging techniques, resulting in multi-angle illumination digital holography, structured light illumination digital holography etc. However, due to the lack of terahertz devices and the strong absorption of terahertz by polar substances, the terahertz full-field phase-contrast imaging usually requires a compact and simple optical path structure. To the best of my knowledge, no relevant work on terahertz image-plane digital holography has been reported yet. In recent years, various terahertz devices such as lasers, detectors and gratings have been developed rapidly. New terahertz imaging elements such as large numerical aperture terahertz lenses and terahertz solid immersion objectives have emerged, providing strong hardware support for the development of terahertz imaging.In this paper, a continuous terahertz wave image-plane digital holographic imaging method based on TPX lens is proposed. The holographic recording and reproduction process is studied, and the quadratic phase distortion introduced by TPX lens is removed by surface fitting distortion correction. The digital holographic imaging system of terahertz image plane is built. The imaging resolution of the system is quantitatively analyzed by using an amplitude Siemens star sample. Then two phase samples of organic materials and biological samples are imaged. The experiment verifies the effectiveness of the phase contrast imaging method. The introduction of image plane imaging in terahertz band is conducive to improve the resolution and quality of terahertz imaging. It has a wide application prospect in biomedical imaging, nondestructive testing and other fields.
We propose an in-plane displacement measurement method based on combination of Digital Holography (DH) and Digital Image Correlation (DIC). It utilizes intensity images reconstructed by DH, and extends the capability of a single DH setup to in-plane displacement measurement.The methdology of the proposed method is illustrated. And the holograms before and after the in-plane displacement of the object are recorded by CCD. Then the intensity images are numerically reconstructed from holograms. And finally the DIC algorithm is applied to the intensity map to obtain the in-plane displacements in u and v components, respectively. The method we proposed solves the problem that one single DH setup can only measure out-of-plane displacement, but not in-plane displacement.Furthermore, the intensity images of DH used in this method is essentially laser speckle images, rather than the surface spray speckle or natural texture of the object under white or natural light used in traditional DIC. Therefore it is a meaningful exploration of DIC method for the type of laser speckle images. Real experiments of in-plane displacement measurements of diffuse planar objects are performed. The results presents an average relative measurement error of 1.17%, verifying the validity and high accuracy of the proposed method.Finally, this paper discusses the feasibility of proposed method for in-plane displacement measurements of diffuse curved objects. Conventionally, 2D-DIC is unable to perform in-plane displacement measurements of curved objects due to the speckle decorrelation phenomenon caused by the limited depth of focus. However, with the ability of the DH technique to focus objects to different depths in a single hologram, it is possible to segment curved objects into different depth parts with each depth part focused for DIC measurements. This paper theoretically discusses such feasibility and experimentally traverses the DH reconstruction distance to evaluate the accuracy of the proposed method for in-plane displacement measurements of curved objects. This is also a meaningful exploration of DIC measurements of in-plane displacement of curved objects with the assist of DH.The advantages of this technique over the existing methods are as follows: 1) The setup is simple and has high information utilization, making full use of the intensity map information that has been discarded in previous holography method, enabling in-plane displacement measurements with only one DH setup; 2) Extend DIC to the type of laser speckle images and prove its effectiveness; 3) With the help of DH capablity in mutlti-depth focusing, the possibility of 2D-DIC for in-plane displacement measurement of curved objects are explored and demonstrated feasible.
The phase shift digital holographic micromeasurement system based on Liquid Crystal Spatial Light Modulator (LCSLM) is simple and easy to operate. However, LCSLM has problems of spatial inhomogeneity and edge field effect, resulting in phase distortion after modulation. To improve the measurement accuracy of phase shift digital holographic microscopy system based on LCSLM, a method of phase shift digital holographic measurement with secondary phase calibration is proposed in this paper. Firstly, based on the Fizeau interferometric system, the relationship between the gray level and the phase shift of the reflection LCSLM was calibrated once, and the phase of the distorted wavefront was measured. Then, two dimensional image interpolation algorithm and grayscale algorithm are used to design distorted phase conjugated grayscale image with the same resolution as LCSLM. Finally, the conjugated grayscale map is loaded on the LCSLM to obtain the ideal modulation phase, and the purpose of the second phase calibration is achieved. The experimental results show that after the second phase calibration, the PV value of LCSLM wavefront difference decreases from 0.182λ to 0.088λ, and the RMS value decreases from 0.039λ to 0.022λ. The overall liquid crystal surface modulation effect is more close to the ideal value, which proves that the proposed method can effectively improve the phase modulation accuracy of LCSLM. The phase shift digital holography micromeasurement system based on LCSLM was constructed, and the measurement experiment of the microlens array was carried out. The results show that under the condition of without using additional noise filtering, calibration before measuring wave of serrated fluctuation, calibration of measuring wave is relatively smooth, the micro lens array longitudinal secondary calibration rise of relative measurement error reduced to 1.15% from 3.08%, which proves that the method can effectively improve the accuracy of measurement of phase shift digital hologram. It can be seen that the precision of phase reconstruction can be controlled by using liquid crystal devices in the interferometry device, and the calibrated LCSLM has strong wave-front control ability, and significantly improved phase images can be obtained. Compared with the traditional mechanical moving phase shift measurement technology, this technology has convenient operation and simple device. Only LCSLM is controlled to change the loading pattern, and data acquisition is faster, reducing the requirements for environmental and other experimental conditions. It can be applied to deformation analysis and 3D measurement of micro-nano devices, so it has good research value and application prospect.
In polarization imaging detection, the difference and change of physical properties of aerosols or detection targets are reflected by polarization characteristics. The high-dimensional polarization characteristics effectively improve the contrast between the target and the background, thereby laying the foundation for realizing the inversion of the target's spatial structure. This feature can enhance the recognition effect of the target in the cluttered background. Affected by the imaging distance and atmospheric interference, the limit resolution of the image projected on the focal plane is greatly reduced (much smaller than the optical system diffraction limit resolution), resulting in a lower spatial resolution of the polarized image. On the other hand, the spatial resolution of the polarized image is limited by the number of detector pixel. High resolution images are of great significance and value to the accuracy of target detection. For this reason, without replacing the hardware imaging system, the super-resolution reconstruction method is usually adopted. This method is a common technical means in image processing and practical engineering applications, and it is also a hot research issue of underlying computer vision. The existing image super-resolution algorithms have some problems, such as low utilization of feature information, large amount of parameters, blurred image reconstruction details and so on. The input features of low-resolution images contain rich low-frequency information, which is treated equally in different channels. In computational imaging process using deep learning, image mapping function solution space is very large, it is difficult to generate detailed texture and high-frequency information lack, which hinders convolutional neural network representation ability in image super-resolution. In order to solve this problem, a depth residual polarization image super-resolution network combined with double attention mechanism is proposed. This paper proposes a dual attention residual network model. The network structure is composed of a residual network with a global jump connection, which realizes the connection between the bottom network and the top network to stabilize the training of the deep network. The residual network contains 10 residual groups, and each residual group contains 20 residual blocks cascaded by dual attention blocks with local skip connections; At the same time, considering the interdependence between channels, an adaptive channel feature adjustment mechanism is designed. The channel attention mechanism is regarded as a guide, which biases the allocation of available processing resources towards the most informative part of the input; Cascaded spatial attention blocks are introduced to focus residuals characteristics on the key spatial contents. Spatial attention function measures the correlation between target features and key features, obtains the attention weight, and then aggregates the key content adaptively. The up-sampling module at the end of the network uses sub-pixel convolutional layers to reconstruct high-resolution images. The experiments mainly consist of two parts: first, use the bicubic degradation model to down-sample the training data set collected by the polarization camera, then add noise and blur to obtain the corresponding low-resolution image, and complete the training of the network model. The proposed method is compared with Bicubic, SRCNN, FSRCNN, EDSR methods to verify the effectiveness of the algorithm. In addition, the image reconstructed by this method is compared with the imaging system to provide data reference for system calibration and correction. The experimental results show that the reconstructed image with this method has richer texture details and uniform brightness, which is closer to the high-definition image of the imaging system. The peak signal-to-noise ratio and structural similarity index of this method are better than other methods, but the amount of parameters is only about 2/5 of that of EDSR. Thus, this method has the advantages of low information redundancy and better reconstruction effect.
Infrared sensors can capture prominent target characteristics by thermal radiation imaging, however the obtained infrared images usually lack structural features and texture details. On the contrary, visible sensors can obtain rich scene information by light reflection imaging, the obtained visible images have high spatial resolution and rich texture details, but cannot effectively perceive target characteristics, especially in low illumination environmental conditions. Infrared and visible image fusion aims to integrate the advantages of the two types of sensors to generate a composite image with better target perception and superior scene representation, which is widely applied for object tracking, object detection and pedestrian re-recognition. The existing generative adversarial network-based fusion methods only make use of convolution operation to extract local features, but do not consider their long-range dependence, which is easy to cause the fusion imbalance, resulting in the fusion image cannot retain typical targets of infrared image and texture details of visible image at the same time. To this end, an end-to-end infrared and visible image fusion method via interactive attention-based generative adversarial network is proposed. Firstly, in the generative network model, we adopt a dual-path encoder architecture with weight parameters sharing to extract the respective multi-scale deep features of source images, where the first normal convolution layer is used to extract low-level features, and two multi-scale aggregation convolution models are adopted to extract high-level features. By aggregating multiple available receptive fields, our multi-scale dual-path encoder network can efficiently extract more meaningful information for fusion tasks without down-sample or up-sample operations. Secondly, in the fusion layer, we design an interactive attention fusion model, which is cascading channel and spatial attention models, to establish the global dependence of their local features from the channel and spatial dimensions. The obtained attention maps can refine multi-scale feature maps to more focus on typical infrared targets and visible texture details, so that the fused results achieve better visual results. Finally, in the adversarial network model, we propose two discriminators, such as Discriminator-IR and Discriminator-VIS, to balance the truth-falsity between fusion image and source images. Besides, we introduce the mutually-compensated loss function to supervise the entire network, which can gradually optimize the generative network model to obtain the best fused result. In the ablation study and verified experiments, the TNO and Roadscene datasets and eight evaluation metrics are proposed to demonstrate the effectiveness and superiority of the proposed method. The ablation experimental results of the interactive attention fusion model indicate that our model can effectively establish the global dependency of local features compared with other four models, and further improve infrared and visible image fusion performance. In addition, compared with other nine the state-of-the-art fusion methods, such as WLS, DenseFuse, IFCNN, SEDRFuse, U2Fusion, PMGI, FusionGAN, GANMcC and RFN-Nest, the proposed method can achieve more balanced fusion results in retaining the typical targets of infrared image and rich texture details of visible image, and has a better visual effect, which is more suitable for the human visual system. Meanwhile, from a multi-index evaluation perspective, the proposed method has better image fusion performance, higher computational efficiency and stronger robustness than other state-of-the-art fusion methods.
The imaging process of the synthetic aperture radar system is not affected by time and weather, and can achieve all-day and all-weather imaging of the target. It has a wide detection range and generates high-resolution images. Therefore, it is widely used in the military and civilian fields. In recent years, satellites of "GF-3" and satellites of "HJ-1" were successively launched to fill the gap in synthetic aperture radar technology in China. However, synthetic aperture radar images still have shortcomings in the target detection process. Ship targets in SAR images are sparse and of various scales. The anchor box-based detection model relies too much on manually designed candidate boxes, which cannot adapt to all ship targets, and the parameters of the candidate boxes consume a lot of computing resources. The background of the synthetic aperture radar image is complex, and the ship target can easily disappear in the complex background, which leads to the missed detection of the detection model. SAR images contain a large number of small-scale ship targets, which are easily lost after multiple convolutions. Coherent speckle noise present in SAR images causes blurring of ship edges. To address the impact of the above problems, this paper proposes a detection model for pixel-level denoising and semantic enhancement. First, the pixel-level denoising module uses the prediction mask to generate the attention map of [0, 1], multiplies the feature map and the attention map pixel by pixel to achieve denoising, and optimizes the attention map using the cross-entropy loss. The denoising module can enhance the weight of the target area of the ship, suppress the weight of the non-target area, and enhance the difference between the ship target information and the background information in the feature map. Second, the semantic enhancement module enhances the semantic information contained in the feature map, and uses asymmetric convolutional layers to extract features of different dimensions, preventing candidate boxes with high IOU scores and low classification confidence from being suppressed. The transformer encoder is introduced in the semantic enhancement module to improve the context information between the ship target and the feature map, and enhance the dependency between the ship target and the image. Finally, the de-noised feature map with rich semantic information is fed into the detection head. In the public data set SSDD test, the model detection accuracy reaches 96.73%, the detection accuracy for small-scale ships reaches 96.85%, the detection accuracy for large-scale ships reaches 96.41%, the detection accuracy for distant sea scenes reaches 98.53%, and the detection accuracy for near-shore scenes reaches 90.00%. Compared with the detection effect of the current mainstream model in the SAR-Ship-Dataset dataset, the proposed model is verified to have a better detection effect. The experimental results show that the pixel-level denoising module uses the attention map to change the weight of the ship target area can better distinguish the target area from the non-target area. In the SAR image with complex background, the ship target position information is more obvious. The position information of small-scale ships is also enhanced, which solves the problem of small-scale ships loss after multi-layer convolution operations. The semantic enhancement module improves the model's ability to recognize ship targets and reduces the suppression of high-quality candidate boxes. Therefore, the model in this paper can effectively reduce the missed detection rate and false detection rate of the model.
Humans mainly perceive and understand the unknown world by obtaining effective information, the visual system has always been an important way to obtain external information. With the development of digital information technology and the demand for human vision, imaging equipment has greatly improved in items of image resolution and dynamic response range. In recent years, imaging technology and its processing technology have played a vital role in various fields. Due to images captured by traditional cameras can only record a limited dynamic range, and the scene is unrepeatable and transient, the interested target cannot be captured again, and we can only process existing images, therefore, reconstructing the high dynamic range image from low-quality images and improving the visual quality of scenes is a key issue in computer vision and has very important research value. In this dissertation, we focus on the dynamic range image reconstruction method in improving image quality of static scenes. The lack of ground-truth fused images for supervised learning, and exiting multi-exposure image fusion suffer from loss of edge features and blurred detail. To address these problems, we propose an attention guided network for multi-exposure image fusion. First, a dual channel Unet network with independent weights is established, extract feature from under-exposure and over-exposure images of the target scene, and a multi-scale and high-dimensional feature maps with strong texture information feature expression ability is obtained. Then, through visual attention mechanism focus local details and global features of under- and over-exposure images, generated the logical mask of the target region of interest area and superimposed on the high-dimensional multiscale feature maps to highlight the target features and suppress the non target area. Finally, during the reconstruction process, we concatenate the filtered high-dimensional multiscale features, the dilation residual dense block is used, the dilation residual dense block makes full use of the features of different levels, retains more detailed information from low dynamic range image, and increases the image receptive filed to predict the details of the saturation region. Based on end-to-end network, in order to reconstruct the fused image more accurately, in which the L2 norm is used as the constraint criterion of the content loss and the SSIM is used as the constraint criterion of the structural loss to design multiple loss functions constrain the neural network, so as to obtain a small similarity difference between the source image sequence and the fused image, realize more accurate convergence of the neural network model, and unsupervised learning. To verify the effectiveness of the proposed algorithm, some images selected from the MEFB benchmark dataset as the test set. The test set including indoor, outdoor, day, night and other static scenes, covering a wide range of real scenes, which can better show the real scene information. Combined three traditional algorithms and two deep learning algorithms for subjective analysis, and used five quality evaluation indicators of fusion image and average running time for objective evaluation. Ablation study were carried out from the effectiveness of both the attention mechanism module and the loss function λ hyper parameters, the experimental results show that the proposed algorithm can capture more detailed information and structural information from the source image sequences under static scenes, obtain fused images with clear scenes and salient features, and the fused image is more in line with human visual characteristics. Comparing with the other typical algorithms, the proposed algorithm not only overcomes the shortcomings of traditional algorithms that cannot adaptively learn features and the fusion rules need to be hand-crafted, but also introduces attention mechanism and dilation residual dense block, which make easier to predict the details and structural information of saturated areas and under-exposure areas, so as to obtain more comprehensive, reliable, abundant scene information with stronger robustness.
With the application of artificial intelligence in agriculture, the requirement of applying machine vision in the field to identify soil species has been raised. Different natural light will bring different images when soil images are collected by machine vision in the field, and it will affect soil species recognition. For refraining from this influence, one method is to collect completely images of soils under a variety of different natural lighting conditions. However, the acquisition of soil images in natural environments can be limited by natural conditions, time and economic costs, and it is difficult to implement. Thus, it may be an effective method that the soil image is converted to be similar to those real soil images that collected in the specific lighting environments, and it can eliminate the influence of inconsistent sunshine environments to improve the accuracy of soil species recognition. The main work of this paper is as follows.Multiple Gaussian fitting of brightness histogram of soil image is realized. Through studying and analyzing soil image, it is found that its Y component histogram is a skewness distribution and its left parts is similar to the left local area of a Gaussian curve, and the remainder that the Y component histogram is fitted by Gaussian curve still remains the features that its left parts is closed to the left local area of a Gaussian curve until the remainder becomes white noise. So the Y component histogram of a soil image can be fitted by several Gaussian curves. Based on the above ideas, an optimization model is established to fit the Gaussian curve of its left local area. Then, the fitting residual is computed and the next Gaussian fitting is executed until the fitting residual is small enough. The weighted fitting curve of multiple Gaussian fitting and weighted Gaussian subtractive fitting algorithm are obtained.Controllable brightness enhancement algorithm of soil image based on weighted Gaussian subtractive fitting is proposed. The target brightness is introduced into the weighted Gaussian subtraction fitting curve of an original soil image to calculate its probability density curve and cumulative distribution curve of the Y component of the expected enhanced image. The brightness migration is raised to realize the controllable brightness enhancement by the cumulative distribution curve of the original soil image and the cumulative distribution curve that the target brightness has poured into it. According to the principle of color ratio invariance, the transferred Y component carries out the color correction of U and V components to get the final controllable brightness enhancement of soil images.Simulation experiments prove that the algorithm can enhance soil image with controllable brightness and is effective. Three sample sets for simulation experiments are constructed and eight experiments are done to test the algorithm. We compare the proposed algorithm with the existing controllable brightness enhancement algorithms, such as 1-D HS and 2-D HS. The experiments include the converting sub-images from high to low brightness and the converting from low to high brightness in each pair of sub-images, and the results indicate that the proposed algorithm has less variation in brightness difference accuracy and color difference accuracy, higher controllable accuracy of brightness and less distortion. Next, experiments on the effective range of brightness adjustment are conducted. The brightness means of a sub-image as its brightness base point, and the step difference is 10 in the brightness increment experiment and the brightness decrement experiment. Simulation results exhibit that increasing or decreasing 30 brightness gray levels is the effective range of soil image brightness enhancement with the algorithm in this paper. Finally, a weighted Gaussian subtraction fitting experiment is performed, the results show that the algorithm can adaptively obtain the fitting times and algorithm convergence is implemented after the iteration is performed 6~8 times, and the fitting accuracy is improved.
The lung lesions segmentation in medical imaging is an important task. However, there are still some challenges. The lesions delineation relies on manual segmentation by experienced clinicians, which is time-consuming and labor-intensive due to the complex anatomical structure of the human body; Lung tumor images have the characteristics of low contrast, different size and shape of the lesions, and variable location of the lesions, and are characterized by unbalanced data distribution. U-Net can segment lesions under a small number of datasets and has been widely used in medical image segmentation of lesions and organs. However, U-Net has the following three problems. First, U-Net uses uniform parameters for each feature map. For lesions of different sizes and complex shapes, the network may have poor spatial perception, which leads to the decline of segmentation performance. Second, U-Net channel dimension doubles with the number of down-sampling, and the feature map of the encoder layer is concatenated to the decoding layer through skip connection. However, in the segmentation task, the importance of different channels to the segmentation task is different. Third, most of the current multi-encoder segmentation networks extract the features of the single-modal target slice and their continuous slices to improve the network segmentation performance, but ignore the ability of different modal medical images to express the characteristics of the lesion. To solve the above problems, this paper proposes the MEAU-Net network to extract complementary features of multi-modals images. First, for the unbalanced data distribution, the Hough transform is used to detect the line of the lung Computed Tomography (CT) image marked by the doctor to obtain the region of interest, and cropped image size from 356 pixel×356 pixel to 50 pixel×50 pixel. Then, for the low contrast of medical image, use exposure fusion image contrast enhancement method improves the contrast between lesion and the background of lung CT image. To extract the features of lesions in multi-modal medical images, this paper proposes a multi-encoder hybrid attention mechanism network MEAU-Net. Positron Emission Tomography (PET) images provide metabolic information of lesions, CT images provide anatomical information of lesions, and Positron Emission Tomography/ Computed Tomography (PET/CT) images combine their advantages and utilize their complementarity and redundancy. MEAU-Net encoder path includes three branches of PET/CT, PET and CT, which are used to extract corresponding modal image features. In the skip connection of the network, hybrid attention mechanism is used, including spatial attention mechanism and channel attention mechanism. The features of PET/CT and CT are used in the spatial attention mechanism to emphasize key areas in the feature map and suppress irrelevant background. The channel attention mechanism extracts the weight value of each channel for the three branches of PET/CT, CT and PET, and then selects the maximum weight value after the three branches sigmoid to multiply the corresponding channel, and assign a higher value to the important channel. The weighting coefficient realizes the selection of important channels. The network inputs the feature map through the hybrid attention mechanism into the corresponding decoder layer, so that the network focuses on the lesion part in medical image, suppresses useless background information, and achieves accurate segmentation of the image lesion. Finally, for the semantic features of different scales of the decoding path, this paper uses a multi-scale feature aggregation block to perform feature mapping on the features of the decoding path, and refine the segmentation of the lesion. We compared our model with 4 classical segmentation model on our dataset, including SegNet, Attention Unet and Wnet. The experiment results show that our model uses multi-modal medical image features to effectively segment lung lesions with complex shapes, and outperforms all other methods in our dataset. The average DSC, Recall, VOE and RVD of MEAU-Net segmentation results are 96.4%, 97.27%, 7.0% and 6.94%, respectively.
Image process is widely used in route planning, industrial damage detection, face recognition, medical aided diagnosis and other fields, and the vigorous development of this technology also has higher requirements for the performance of image acquisition equipment. The insufficient exposure and inconspicuous texture of the detected object in the image will be affected by the general deviation of the image quality collected by the low-end equipment, while the high-end image acquisition equipment is generally expensive and has a precise structure, which is not suitable for use in harsh acquisition environments. At the same time, the distance between the position where the detected object appears and the image sensor is also affected by uncontrollable factors such as randomness. These bad factors all increase the difficulty of the later target recognition task, and also bring a bad visual experience to the user. Therefore, it is of great theoretical significance and application value to design an enhancement algorithm to improve the quality of low-light images.To solve the problems of low contrast and blurred details in low-illumination images, a multi-domain block low-illumination image enhancement algorithm fused with genetic algorithm is proposed. The algorithm can be divided into four stages: color space conversion, brightness enhancement, detail enhancement and multi-scale fusion. First of all, to prevent the original color characteristics of the image from changing when the image brightness is enhanced, the input image is converted from the RGB color space to the HSV color space, so as to better separate the color information and brightness information of the input image, so that the color of the image can be improved. Information is not altered when augmented. When the algorithm enhances the image brightness, in order to prevent the over-exposure or under-exposure of some areas of the processed image, it is necessary to reduce the impact of the complex exposure of the acquisition scene on the enhanced image brightness. Therefore, the multi-threshold block enhancement is more in line with the actual scene. The method is used to enhance the brightness of the image. In order to improve the processing speed of the algorithm, the genetic algorithm is used to search for the optimal segmentation threshold of the brightness component of the input image. Then, the luminance channel of the input image is divided into a plurality of different exposure level sub-images according to the obtained multiple thresholds and the image luminance gradient law. The detailed information contained in each sub-image is the evaluation criterion, and the complexity of all sub-images is evaluated through a multi-threshold block enhancement algorithm. The brightness of each sub-image is adjusted according to the evaluation results, and the arrangement order of the brightness of each sub-image after enhancement same as that before enhancement. To enhance the detail information of the image, the guided filtering algorithm is introduced, and the original image is subjected to two guided filtering processes. The first process filters out image noise, and the second filter enhances the contour information of the image. After two filtering processes, the enhancement results with rich contour information and less noise are obtained. The unsharp mask algorithm is introduced, which uses low-pass filtering to obtain the low-frequency information of the original image, subtracts the original image and the low-frequency information to obtain the high-frequency information of the image, and superimposes the enhanced high-frequency information with the original image to obtain the high-frequency information of the image. Finally, a multi-scale fusion algorithm is introduced to decompose the enhanced input image contour information, enhanced input image texture information details and brightness enhanced input image into Laplacian pyramid and Gaussian pyramid. The resulting Laplacian input and the corresponding Gaussian weight map at each level obtain the final enhancement result from texture information, contour information and brightness information.The algorithm is also compared with the existing enhanced algorithms. The results show that the increase of each index of the image enhanced by the proposed algorithm is greater than that of other comparison algorithms, and the proposed algorithm effectively solves the problem of color distortion and brightness blocking while enhancing image brightness, effectively restoring the texture information of the image. At the same time, the brightness distribution of the enhanced image restores the brightness of the real shooting environment. It proves that the algorithm has better performance.
The composite nanostructures composed of different semiconductors endow them with the high separation rate of photogenerated electrons and holes through the charge transfer effect at the heterojunction interface, to achieve high performances of the photocatalysts. Based on such nanostructure, multiple nano-heterostructure photocatalysts has become a research hotspot in the field of photocatalysis. Among them, ZnO/TiO2 nanocomposite material exhibits excellent photocatalytic performances, while the existing synthesis process is complex and limits its industrial application in photocatalysis. For this reason, based on the modified polymer network gel method, we prepared a highly efficient ZnO/TiO2 nanocomposite photocatalyst by the simply designed process routes, and adjusted the ratio of the components in the composites to optimize the photocatalytic performance. The TiO2 content was calculated according to the molar ratio of Ti/(Ti + Zn), which are 0.2 mol/mol, 2 mol/mol, 10 mol/mol, 25 mol/mol and 50 mol/mol respectively. The corresponding samples were labeled as ZT0.002, ZT0.02, ZT0.1, ZT0.25 and ZT0.5. The phase compositions, morphology features, microstructures, surface chemical states, and optical and electrical properties of the catalysts were characterized by X-ray diffraction, scanning electron microscope, transmission electron microscope, inductively coupled plasma mass spectrum, X-ray photoelectron spectroscopy, electron paramagnetic resonance spectrum, ultraviolet and visible spectrophotometry and steady-state surface photovoltage spectra. Also, we studied the photocatalytic degradation characteristics of the catalysts toward methylene blue and methyl orange under simulated sunlight irradiation. Correspondingly, the mechanism about the change in the composition of the composite photocatalyst causing the change in its catalytic activity was proposed.The results of the study indicated that the addition of a small amount of TiO2 and compositing with the high-concentration TiO2 both improved the catalytic activity of the granular ZnO nanophotocatalyst. Under simulated sunlight irradiation, they possess the higher degradation efficiencies to methyl orange, and the enhanced performances are respectively attributed to the enhanced charge carrier separation efficiencies caused by the increased surface oxygen vacancy defects and the enhanced interfacial charge transfer in the multi-heterojunction structure. To be specific, the phase composition and microstructure analysis indicated that the only hexagonal wurtzite phase ZnO is observed in ZT0.002, while there is a multi-heterojunction structure composed of hexagonal wurtzite phase ZnO, anatase phase TiO2 and cubic phase Zn2TiO4 for the samples composited with the high-concentration TiO2. The morphology feature analysis indicated that ZT0.002 and ZT0.5 exhibit less particle agglomeration, uniform particle size distribution and larger average particle size compared with other samples. Although the large average particle size results in a small specific surface area, less particle agglomeration reduces the recombination of electrons and holes migrating to the interfaces between the crystals. The surface chemical state and defect situation analysis indicated that ZT0.002 has a higher concentration of surface oxygen vacancy defects than that of ZnO, which is caused by the regulation of the nucleation, crystallization and crystal growth process of ZnO by a trace amount of TiO2. The increased surface oxygen vacancies promote the separation of photogenerated carriers and increase the active adsorption sites for the reactant molecules. Likewise, the photogenerated carrier separation efficiency is enhanced by the formation of the multi-heterojunction structure among ZnO, TiO2 and Zn2TiO4.In addition, ZnO and TiO2 showed a certain selectivity for the photocatalytic degradation of methylene blue and methyl orange, which is related to the electrification of the catalyst surface and the ionicity of pollutant molecules. Radical scavenger experiments showed that the surface active species such as hydroxyl radicals and superoxide radical as well as photogenerated holes participate in the degradation reactions of methylene blue and methyl orange on the catalyst surfaces, and superoxide radical plays a dominant role in the degradation of methyl orange. Cycling experiments evidenced the high stability of the as-prepared photocatalyst. For practical applications, the influences of the catalyst dosage and pH value of the pollutant solution on the photocatalytic performance of the catalyst have also been studied.
Two-dimensional (2D) layered materials with excellent properties such as strong spin-valley coupling, high current on/off ratio and carrier mobility, have gained great attentions in optoelectronic applications including photodetectors, field effect transistors and light-emitting diodes. Meanwhile, many Novel Nonlinear Optical (NLO) phenomena such as saturable absorption, reverse saturable absorption and two-photon absorption have also been observed. Based on these nonlinear absorption behaviors, ultrafast photonic applications including Q-switching, mode locking and optical limiting have successfully been designed in 2D materials. The device performance is strongly dependent on their thicknesses due to tunable electronic properties with layer. For example, giant two-photon absorption can be observed in monolayer WS2 film with 1.9 eV while excellent saturable absorption has been demonstrated in multilayer. However, the related research about NLO nonparametric processes with layer thickness, which is crucial to design high performance photonic devices, was seldom reported due to uncontrollable thickness in large-area. Therefore, seeking for an efficient way to produce large-scale and controllable film is highly desirable. To date, mechanical exfoliation have been developed to acquire few-layer nanosheet through exfoliating bulk material layer by layer. However, these methods are low reproducibility and low yield and even limit their application in photonic devices. Compared with mechanical exfoliation method, liquid phase exfoliation was quickly developed and most 2D materials were successfully exfoliated to few-layer or monolayer nanosheet. The nanosheet dispersion can be used to form wafer-scale film by vacuum filtration technique, which provides an effective way to prepare large-area film. Herein, WS2 nanosheet dispersion was prepared by liquid phase exfoliation and then WS2 films with different thicknesses were prepared by a vacuum filtration method with different volumes. The atomic force microscopy confirmed the nanosheet length is between 2 and 4 μm and the thickness of thin film is 50 nm for 30 mL in fltration volumes. Raman spectroscopy was used to charaterize the sample composition and the typical Raman peaks confirmed the preparation of WS2 films. UV-Vis spectra was used to obtain the linear absorption and vetified the defects in the grain boundary. The third-order NLO absorption of prepared WS2 nanofilms was studied by a Z-scan technique based on 800 nm femtosecond laser. It is found that all WS2 with different thicknesses show the characteristics of saturable absorption, which is mainly due to the Pauli blocking effect caused by single photon absorption. To confirm the absorption process, the electronic structure was calculated by density functional theory and the calculated bandgap of bulk WS2 is approximately 0.8 eV. The photon energy is higher than the bandgap, which vertified the single photon absorption occurs. The nonlinear absorption coefficients are -150 cm/GW, -139 cm/GW, -133 cm/GW, -122 cm/GW, and -115 cm/GW from 30 mL to 70 mL with per 10 mL fltration volumes. Obviously, the absolute value of nonlinear absorption coefficients decreases with the incereased thickness. This decrease is mainly due to the reason that thicker films with high density defects capture more photoexcited carriers and generate more nonlinear scattering as well as energe loss, resulting in the dependence of third-order nonlinear optical absorption with thicknesses. The existence of certain defects will introduce defect energy levels, which has a great potential in the application of infrared and near-infrared Q-switched and mode-locked lasers. With the increase of thickness, the saturation intensity is basically unchanged, and the modulation depth increases while the absolute values of the imaginary part of the third-order nonlinear optical susceptibility and figure of merit decrease. Our results provide provide experimental support for the applications of WS2 thin films in ultrafast switchers and ultrafast lasers.
Titanium alloy is a kind of metal material which with a wide range of applications, including aerospace, rail transit, medical equipment and other fields. The appearance of titanium alloys with different brands is very similar, but they are suitable for different fields. Even the properties of titanium alloys with the same brand and different numbers are different, and confusion is easy to cause serious accidents. Therefore, it is urgent to study the rapid and accurate classification and identification of titanium alloys with the same brand. In recent years, laser-induced breakdown spectroscopy as a fast, real-time, in-situ, micro-loss, multi-element synchronous analysis of advanced detection technology, favored by researchers. Using the advantages of laser-induced breakdown spectroscopy technology, the titanium alloys with different national standard numbers under the same brand are analyzed, which can realize the rapid and accurate classification and identification of titanium alloys.The whole device used in laser-induced breakdown spectroscopy is composed of laser, spectrometer, electronic control displacement platform, workstation, acquisition head, digital delay generator and several lenses and optical fibers. According to sequential analysis, the spectra of titanium alloy under various laser intensities and different delay times were collected by the device. Combined with previous literature research and experience, six characteristic spectra with high signal strength and high signal-to-noise ratio were selected. The optimal laser intensity and trigger delay were obtained by comparing the peak intensity and signal-to-noise ratio of the six characteristics. The TC4 titanium alloy spectrum collected under the optimal conditions is divided into training set and test set according to the ratio of 3∶1. The training set data is trained by a variety of algorithms, and then the test set is substituted into the trained model. Through the analysis of the two results, the advantages and disadvantages of the algorithm are determined, and the optimal algorithm is K-nearest Neighbor algorithm. The optimization of data is mainly carried out from three aspects: 1) By using 3σ method for data screening at 10 characteristic spectra, the inferior spectra with too large or too small peak values are eliminated, to avoid its impact on the results; 2) Through data normalization, reduce the impact of experimental environment and experimental parameters; 3) Through principal component analysis, the data dimension is reduced, a large number of redundant data is reduced, the classification accuracy is improved and the model training time is reduced. KNN model can optimize the parameters mainly include the number of adjacent points, distance measurement, and distance weight. The number of adjacent points is the core parameter of KNN, which determines the number of data used to determine the unknown points. Distance measure is the distance calculation method between two points. Under different distance measures, the data between two points are different. Distance weight is the relative importance of determining the distance between the known point and the unknown point. The three parameters are arranged and combined into the model to retrain, and the optimal parameter combination is determined by comparing the results of the training set and the test set.The optimal classification results are obtained through various work, and the classification and recognition of the same grade titanium alloy are realized. The results show that the classification accuracy of the same grade titanium alloy can be improved from 84.15% to 99.14% by combining data processing and model optimization, and the training time can be reduced from 1232.41 s to 83.91 s. The classification performance is significantly improved. The research results are expected to achieve rapid and accurate classification and identification of titanium alloys with the same brand, and have broad application prospects.
Murals are treasures in the long history of Chinese culture. It has high research value in history, science and art. Pigment is the material carrier of the main form of mural expression. Simultaneously, it is also an important part of murals. After a long period of disease and corrosion, the surface of the mural may be damaged to varying degrees, making it difficult for researchers to distinguish the type of pigments in the murals. The accurate identification of pigments is the premise of conservation and restoration of cultural relics. The traditional method needs to take samples from the murals, which will cause irreversible damage to the murals. In this paper, multispectral imaging technology and deep learning related classification algorithm are used to analyze and identify the pigments in mural multi-spectral images to assist researchers in mural identification and cultural relic restoration. Rich spatial and spectral information is included in mural multispectral images. In traditional algorithms, spatial or spectral information is used as a feature of mural multispectral image classification. This method leads to low classification accuracy of mural multi-spectral image. In order to improve the classification accuracy of mural multispectral images, the deep learning algorithm is used in this paper, which can make full use of the spatial and spectral information of multi-spectral images. In the actual shooting, due to the limitations of site conditions and protection requirements, mural spectral imaging data need to be collected quickly. The efficiency of data acquisition can be improved by using sparse channel imaging methods. However, this method can make the spectral reflectance curve of pigments appear nonlinear, which can affect the classification accuracy of mural multi-spectral image. In order to solve this problem, a pigment classification method for mural sparse multi-spectral images based on spatial spectral combination features is proposed in this paper. The nonlinear spectral features are extracted by using the hyperbolic tangent activation function in the Long Short-term Memory (LSTM). Firstly, the spectral reconstruction of the mural multi-spectral image is carried out. Then the one-dimensional spectral vector is input into LSTM, which can actively learn under unsupervised conditions to reduce the influence of spectral curve nonlinearity on the classification accuracy. In order to solve the problems of high spatial resolution and strong correlation between adjacent pixels in multispectral imaging, the linear correction function in Convolution Neural Network (CNN) is used to map the feature map to nonlinear space. The activation function is added after the convolution operation of each layer to improve the nonlinear expression ability of the mural multi-spectral image. The spatial spectral unity of mural multi-spectral images can not be fully utilized, if only spatial or spectral features are used. For this reason, the multi-scale fusion strategy combining spectral and spatial features is used to eliminate the influence of spectral nonlinearity and spatial correlation on the classification results. Firstly, a Spatial Spectral Joint Feature Network Model (SSJF) is established to train pigment samples. Then, the loss function is designed by cross entropy, and the gradient is updated by back propagation algorithm. Finally, the softmax classifier is used to output the probability of each pigment. The experimental results show that the pigments in the paint board and self-made murals can be correctly classified through SSJF. The OA and Kappa coefficients reached 97% and 0.97, respectively, which effectively improved the pigment classification accuracy of the sparse multi-spectral image of the mural.
As an important driving factor of climate change of the global, water vapor plays an important role in the material and energy transmission and evolution of the earth system. As a natural greenhouse gas, the particularity of water vapor is that it can change among gas, liquid and solid under natural conditions, resulting in drastic changes in its abundance. With the intensification of global warming, the water vapor content in the atmosphere will also increase, thus forming positive feedback and accelerating warming. Isotope tracing is an important means to study the atmospheric cycle. The scientific problem of water vapor source and sink can be solved by obtaining water vapor isotope information. The distribution of water vapor is affected by many factors (such as surface geographical environment, latitude, temperature, etc.), and the concentration of water vapor varies greatly in different places and at different times. This requires that the measuring instrument must have the characteristics of miniaturization and portability in the technology to ensure the performance. As an atmospheric remote sensing method developed rapidly in the world in recent years, laser heterodyne spectroscopy has the characteristics of high signal-to-noise ratio, high spectral resolution and small volume. Based on this, using laser heterodyne technology to obtain the whole layer of atmospheric water vapor isotope information is an effective means. A system prototype of all fiber laser heterodyne spectrometer is established by using a self-developed high-precision solar tracker. In this paper, the laser heterodyne spectroscopy method is used to detect the atmospheric water vapor isotope HDO through a set of laser heterodyne device whose central wavelength is located in the water vapor isotope HDO (1 553 cm-1). The field atmospheric simulation and detection were carried out in Hefei Science Island area. The laser heterodyne spectral signal of water isotope HDO molecules was obtained in the near-infrared band (6 437~6 441cm-1). The laser heterodyne device takes the sunlight as the signal light, scans the laser with the wavelength at the absorption peak of water vapor isotope HDO as the local oscillator, obtains the high-resolution heterodyne spectrum signal of sunlight and laser, and obtains the whole atmospheric transmittance spectrum of HDO through wavelength calibration and dispersion standardization. The spectral resolution and signal-to-noise ratio of the heterodyne system are calculated to be 0.019 6 cm-1 and 46 respectively. When the absorption line is very low, it still shows good spectral resolution and signal-to-noise ratio. At the same time, the inversion of HDO gas concentration is carried out, and the vertical profile of troposphere is obtained.The measurement practice shows that the laser heterodyne technology can be used to study the detection technology of atmospheric water vapor isotope HDO, which provides a new means and method for the detection of atmospheric water vapor isotope HDO.
The related petroleum products need to be upgraded using various microbial degradation, thermal and chemical methods to avoid potential hydrocarbon contamination. Although these techniques are commonly used, they have several limitations, such as high energy consumption, high cost, poor efficiency, and environmental pollution. To date, considerable attention has been paid to the degradation of related petroleum products by ultrasonic and magnetic methods. Ultrasound methods have the advantages of easy operation, mild conditions, high efficiency, low cost, and environmental friendliness. Comparatively, some authors have shown that magnetic and electric methods may improve the rheological properties of crude oil. Apart from this study, to our knowledge, no work studying the effects of the ultrasonic or magnetic treatment of Lube Base Oil (LBO) has been performed to date, and the interaction between the ultrasonic/magnetic fields and organic compounds needs further study. The use of these physical fields to treat and degrade petroleum products has sparked considerable controversy. Terahertz Time-Domain Spectroscopy (THz-TDS) has recently received considerable attention with the expectation that it will provide new insights into complex petroleum systems, such as optical property and spectroscopic studies of the selected lubricating oil and probes of the disaggregation of crude oil. In this paper, the degradation mechanism and molecular dynamics of LBO subjected to annealing, ultrasonic and magnetic treatments were compared and investigated using THz-TDS. The thermal degradation of the Korea105 was assessed by means of an annealing treatment at 110 °C. The ultrasonication times for the samples were 4 (us4h), 6 (us6h), and 12 hours (us12h). The magnetic treatment times were controlled at 4 (mag4h), 6 (mag6h), and 12 hours (mag12h). By comparison with reference and sample pulse, and use of a numerical fast Fourier transform, the refractive index n(ω) and the absorption coefficient α(ω) of the samples were calculated. The change of refractive index of LBO under ultrasonic field is higher than that under annealing and magnetic treatment, owing to the high temperature and pressure rise due to ultrasound, whereas heat interacts on longer timescales at lower energies. Moreover, the absorption peaks of the original Korea105 at 1.7 THz and 2.3 THz corresponding to intermolecular interactions were selected as marker bands to illustrate the interaction mechanism. The absorption peaks of samples varied significantly after annealing, ultrasonic and magnetic treatments. The sonication of Korea105 proceeds by three categories of reactions, i.e., free-radical generation, propagation, and termination. The principle sonochemical process in LBO appears to be C-C bond cleavage and the recombination of alkanes and aromatics, according to the disappearance of or significant shift in the absorption peaks at the two positions corresponding to the intermolecular and parallel alkane interaction forces for all the LBOs after ultrasonic treatment. The effect of a magnetic field on the viscosity of related petroleum oil is very controversial and the controversy was simplified to two aspects. Several researchers have determined that the viscosity of paraffin-based crude oil increases after exposure to a strong magnetic field at 1.5 T for an extended time. In contrast, other studies have found that the magnetic field reduced the viscosity. Our experiments regarding the increasing viscosity of Korea105 verified the first aspect. The increase in the viscosity of Korea105 suggests that the strong magnetic field and long interaction time induced the aggregation and orientation of the organic molecules. The results obtained in this work demonstrate that THz-TDS is a powerful tool that can provide evidence for the degradation mechanism of organic molecules subjected to different physical treatment methods.
The optical fiber sensing system is widely used in many fields, such as long-distance oil and gas pipelines, tunnel safety detection, large structural safety detection, perimeter security detection and so on. Optical fiber sensing signal identification plays a key role in real-time monitoring, abnormal alarm and other aspects. Its working performance directly determines the performance of the real-time, accuracy and stability of the optical fiber sensing detection system. Therefore, fast and accurate identification and classification is of great significance to ensure the safety of various fields and reduce the cost loss caused by equipment damage.To improve the real-time and accuracy of optical fiber vibration signal pattern recognition, a feature extraction algorithm based on compensation distance estimation is proposed. The algorithm draws on the human auditory perception mechanism and extract the MEL frequency cepstrum coefficients from the optical fiber sensing vibration signals. The algrithm uses the compensation distance estimation technology to formulate the feature selection strategy,and finally realizes feature evaluation and optimization. The MFCC feature extraction algrithm can extract the feature of the vibration signal acquired by the optical fiber sensing system, and then identifies the interference signal according to the modal prediction.However, the extracted feature vector by the MFCC feature extraction algorithm has the problems of high dimension and vector redundancy . When these feature vectors are trained and recognized by the classifier, it will increase the time cost and reduce the recognition accuracy. Therefore, effectively reduce the dimension of the MFCC eigenvector is the key to improving the real-time performance and accuracy of optical fiber sensing vibration signals.This paper proposes a feature extraction method based on compensated distance estimation. The CDET algorithm jointly evaluates the intra-class and inter-class discreteness of eigenvectors. The feature evaluation is performed on different dimensions of the feature matrix, and the redundant vector of low score is deleted from the original feature vector matrix, thereby realizing feature dimension reduction. Solve the influence of redundant vectors on classification, avoid the problem of complex operation caused by too many extracted feature dimensions, and improve real-time performance.The steps of the algorithm are to calculate the average distance between samples of the same condition and different conditions , then average the within-class and between-class distances, and then calculate the within-class variance and between-class variance factors. Then calculate the compensation coefficient between the two variance factors, calculate the ratio of the inter-class distance to the intra-class distance, multiply the compensation coefficient to obtain the distance evaluation standard, and select a better feature dimension according to the distance evaluation standard. The experimental results show that the feature extraction algorithm of vibration signal based on compensation distance estimation technology can effectively reduce the redundant vectors that affect the classification accuracy in optical fiber sensing system. It solves the problems of feature representation and operation complexity for the vibration signal, and further improves the effectiveness and real-time performance of vibration signal pattern recognition of the optical fiber sensor system. Compared with Principal Component Analysis (PCA), our algorithm has the same performance in the low-dimensional case. With the increase of dimension, the performance of the algorithm in this paper is better in recognition accuracy. In terms of anti-noise performance, in the presence of noise, in order to improve the feature identification, the dimension of the feature vector extracted from the MFCC feature increases. The PCA method is difficult to distinguish the features caused by noise, and the algorithm in this paper can reduce the influence of superimposed noise by pruning redundant vectors, and can extract feature vectors with high feature recognition. Therefore, the proposed algorithm also has certain anti-noise performance.
The traditional polarization-maintaining fiber alignment method is divided into longitudinal and lateral observation methods. However, both of them have severe defects. For example, the longitudinal observation method limits the application scenarios due to the destruction of the polarization fiber in the detection section. In contrast, the widely used lateral observation method requires high standards for experimental equipment and complex manual operations and alignment algorithms. Regarding the panda polarization maintaining fiber as the research object, instead of the traditional polarization-maintaining fiber alignment thinking of applying a high standard light source and adjusting the imaging surface. Based on applying the matching liquid mixed with glycerin and deionized water, the ideal refraction environment can be created due to the reduced effects of multiple spherical aberrations. Hence, high accuracy visualization can be achieved by projection from the incoherent parallel light sources. Then, applying the rotating motor with a fiber clamp setup and imaging system to form an experimental device, a long-distance section of the polarization-maintaining fiber can be aligned. Additionally, the LabVIEW platform can demonstrate the results of direct observation and real-time extraction of light intensity distribution information, which contains angle information of the polarization axis from the polarization-maintaining fiber. This information can also be called the characteristic value of the high accuracy alignment method, i.e., the three segment diffraction fringe spaces d2, d3, d4, which are sensitive to the polarization axis angle θ inside the polarization-maintaining fiber. According to the internal structure of the polarization-maintaining fiber and the geometric characteristics of the eigenvalue, the geometric relationship between the eigenvalue and the angle of the polarization axis θ of the polarization-maintaining fiber can be deduced theoretically. Furthermore, through computer processing of the experimental data and theoretical analysis of the light intensity distribution, the relationship between the characteristic areas and the angle of the polarization axis was obtained. Thus, real-time alignment between the polarization axis of the polarization fiber and the observation surface could be realized. In the case of the image processing algorithm, the collected information is primarily binarized and converted into gray-scale digital signals, and then the collected data are normalized to improve the signal-to-noise ratio and reconstruct the image edge by the two-dimensional wavelet algorithm. Due to these algorithm techniques, the accuracy of the alignment method has been improved significantly, the deviation ratio is better than 1.32%, and the accuracy of the alignment method is up to ±0.3°. Moreover, polarization-maintaining fibers might have several process defects, and the primary influencing factors are material purity and environmental impacts during the drawing manufacturing process. All of these defects lead to three typical asymmetries of the panda eyes distribution: the radii of the two panda eyes are inconsistent, the distance between the center of the two panda eyes is inconsistent, and the line of the center of the two panda eyes is not collinear with the line of the center of the fiber core. According to the analysis of geometric theory, the high-precision positioning method is also compatible with the errors caused by the asymmetry of the panda eyes distribution.
With the development of science and technology, people have higher and higher requirements for the capacity of network communication. Conventional single-mode single-core optical fibers have gradually approached the Shannon transmission limit of 100 Tbit/s. mULTICORE FIBER (MCF) based on Space Division Multiplexing (SDM) technology has good application prospects for the problem of upcoming capacity shrinkage. However, a specific issue related to MCFs is Inter-Core Crosstalk (ICXT) which can degrade signal quality and limit SDM performance significantly. In real MCF, due to the random perturbations and additional birefringence fluctuations, the ICXT changes randomly in the longitudinal direction. Therefore, approaches to estimate the ICXT are required to analyze the performance of an MCF transmission system. Coupled Mode Theory (CMT) and Coupled Power Theory (CPT) are common methods for studying the coupling characteristics of weakly coupled MCFs. In this paper, the CMT and CPT are optimized considering the influence of stochastic bending and twisting perturbations in the actual fiber laying process, and the ICXT estimation expression is obtained. For the CMT, due to the consideration of bending and twisting perturbations, the equivalent propagation constant in the Coupled Mode Equation (CME) is not constant, but the relationship related to the transmission distance, bending radius, and twisting rate. Therefore, it is impossible to obtain analytical expressions by directly solving the CMEs. The MCF can be divided into N segments by the segmentation method. In each segment, the distance is small enough so that its equivalent propagation constant and the incident power can be approximately regarded as a constant in weakly coupled MCFs. The final ICXT result is obtained by superimposing N sections of ICXT which is obtained by solving the CME in a short section. We call this model the optimized coupled mode theory model (OCMM). For the CPT, the local power coupled coefficient can be obtained by defining the equivalent propagation constant which contains the influence of bending and twisting perturbations. And the optimized average power coupled coefficient can be obtained by averaging the twisting rate. Finally, the crosstalk estimation expression can be obtained by substituting the optimized power coupled coefficient into the coupled power equation and solving it. We call this model the optimized Coupled Power Theory Model (OCPM). Simulations are carried out to compare the above two optimized models with discrete change models and experimental data. The simulation and experimental data are in good agreement. The OCMM results that ICXT as a function of the MCF length show an upward trend of oscillation along the results of the Discrete Change Model (DCM), in which the crosstalk cumulatively increases near the phase matching point and is almost unchanged near the non-phase matching point. Therefore, the OCMM can better reflect the fluctuation of the physical structure of the fiber, while the OCPM shows linear average crosstalk. In addition, the ICXT as a function of the bending radius in the actual homogeneous and actual inhomogeneous MCF is simulated, and the results of the two models are in good agreement. However, in the actual homogeneous MCF, the simulation results of OCPM and OCMM will be somewhat different. This is because OCPM is a simulated average crosstalk, so OCMM is more accurate in this case. In the actual non-homogeneous MCF, the simulation results of OCPM and OCMM are in good agreement. In this case, the calculation speed of OCPM will be faster. In summary, this paper derives optimized crosstalk estimation models based on CMT and CPT, respectively, and verifies the correctness and accuracy of the theoretical model through comparative studies of simulation and experiment. We can select suitable theoretical models for different actual situations.
The ocean is abundant in chemical and power resources, it is the space for human survival and development. Underwater Wireless Communication (UWC) technology realizes the wireless transmission of ocean exploration information, and has received extensive attention from researchers in recent years. Currently, radio frequency communication technology and acoustic communication technology are two comparatively mature technologies in seawater communications, in which the problems conceal. In the radio frequency communication, the large volume of transceiver, the high cost and energy consumption, and the rapidly attenuated radio wave underwater would make it impossible to achieve long-distance transmission and high-speed underwater communication. As for the acoustic communication technology, its large-size equipment, high power consumption, low transmission rate, limited available bandwidth, and severe multipath effects during transmission, would cause speed limits of the data transmission and the increase in bit error rate. With the advantages of no electromagnetic radiation, fast speed, strong mobility, good safety, high bandwidth and green environmental protection, the underwater wireless optical communication has become a new choice for underwater sensor data transmission and acquisition of marine monitoring information, thus playing a paramount role in the detection in underwater environments and development of marine resources. Specifically, the lower loss of blue-green light caused by seawater absorption and scattering can help to reach the underwater transmission rate as Gbit/s. Therefore, blue-green light is used to carry information to realize long-distance underwater transmission. However, as the light beam propagating in seawater, not only will it be affected by the attenuation effect of the absorption and scattering of the seawater impurities, but also easily affected by ocean turbulence caused by fluctuations in refractive index. Ocean turbulence will lead to a series of problems, including light intensity flicker, beam drift, beam expansion, wavefront distortion and other effects, turning to the consequence of the optical signal fading, degrading the communication quality and deteriorating its performance.In previous studies, turbulence was regarded as isotropic, in which the vortex structure (that is, the spatial frequency of turbulence) was symmetric in different directions, and isotropic turbulence was a simple and idealized model. The ocean turbulence occurring naturally is often anisotropic. Ocean turbulence is composed of eddy structures with different sizes and frequencies (that is, eddy structures are asymmetric in different directions). Therefore, this paper considers the asymmetry of ocean vortex motion (that is, the case where the horizontal scale of the vortex is much larger than the vertical scale), as well as uses ocean turbulence parameters and anisotropy factors to express the equivalent structural parameters of ocean turbulence, applying the equivalent structural parameters of ocean turbulence expressed by ocean turbulence parameters and anisotropy factors, the extended Huygens-Fresnel principle and the asymptotic Rytov theory are also used to derive the average received optical power and radiation of a finite-size detector. Meanwhile, the paper also conducts research on the degree of flux variance, and the packet error rate performance of the double-headed pulse interval modulated Gaussian beam in anisotropic ocean turbulence under the Gamma-Gamma turbulence channel model. DHPIM has its built-in symbol synchronization and slot synchronization functions. Compared with PPM and DPIM, DHPIM has shorter symbol length, higher transmission rate, larger transmission capacity, higher bandwidth requirements and better ability to resist multipath dispersion. In the simulation analysis of ocean turbulence parameters (temperature variance dissipation rate, turbulence energy consumption) under different anisotropic ocean turbulence Spread rate, the ratio of the contribution of temperature and salinity fluctuations to the power spectrum), bit resolution and transmission rate, photodetector responsivity and the impact of link distance on the packet error rate, the results come out and indicate that: no matter what kind of ocean turbulence parameters, the performance of the wireless optical communication system is all proportional to the anisotropy factors in the x-direction and y-direction. For different anisotropy factors, when the temperature variance dissipation rate χT decreases, the ratio of temperature and salinity contribution to the power spectrum ω decreases, or the turbulent energy dissipation rate εincreases, the intensity of ocean turbulence can be weakened, and the system decreases accordingly. For the same bit error rate level in the isotropic ocean turbulence, the wireless optical communication system in the anisotropic ocean turbulence can achieve a longer transmission distance. Simultaneously, approaches such as reducing the bit rate, increasing the responsivity of the photodetector, applying a smaller modulation order while appropriately using a larger diameter aperture to receive the average signal fluctuations, would resist the impact of ocean turbulence and reduce the system's packet error rate effectively . This study provides a certain reference value for improving the performance of underwater wireless optical communication systems in an anisotropic ocean turbulent environment.
Optically pumped spin-polarized vertical-cavity surface-emitting lasers might provide properties superior to electrically pumped vertical-cavity surface-emitting lasers, such as faster modulation dynamics, lager modulation bandwidth, lower threshold, and stronger polarization determination. New applications of optically pumped spin-vertical-cavity surface-emitting lasers are foreseen in high-speed optical communication, optical information processing, data storage, quantum computing and biochemical sensing. Various forms of ultrafast instability are observed in optically pumped spin-vertical-cavity surface-emitting lasers, including periodic oscillations, polarization switching, and chaos dynamics. Due to weak material and cavity anisotropy, the output of electrically pumped vertical-cavity surface-emitting laser usually includes two orthogonal polarization components, which is beneficial to realize dual channel optical communication. Therefore, the dual-channel secure communications based on electrically pumped vertical-cavity surface-emitting lasers have received extensive attention in recent years. In most of these examples, the drive-response electrically-pumped vertical-cavity surface-emitting lasers system was always used for secure communications. In such a system, chaotic x-polarization component and y-polarization component were yielded via the introduction of the external perturbations, typically feedback in the drive vertical-cavity surface-emitting laser. The response vertical-cavity surface-emitting laser yielded similar chaotic x-polarization component and y-polarization component when its parameters were identical to those of the drive vertical-cavity surface-emitting laser. Moreover, chaos synchronization between each pair of polarization components played a key role in security and encrypted message recovery. However, the realization of high-quality chaos synchronizations relies on the assumptions that the drive and response electrically pumped vertical-cavity surface-emitting lasers are completely symmetrical in structure, and their parameters match perfectly. In addition, previous works showed that due to the existence of two polarization components in drive and response vertical-cavity surface-emitting lasers, the structural symmetry of these two lasers is broken, which leads to the degradation of the quality of chaos synchronization. Under this asymmetric structure, high-quality chaos synchronization can be received by limiting a certain delay difference between the self-feedback delay of the drive vertical-cavity surface-emitting laser and the channel delay. The assumptions and limits as described above do not hold in practice. Realization of high-quality chaos synchronization will meet much more challenges in practice due to the inevitable imperfect match between the driving and response vertical-cavity surface-emitting lasers, and the variation of the delay difference at any time.Optically-pumped spin-vertical-cavity surface-emitting laser has better controllability for polarization switch, which is conducive to the realization of two parallel reservoir computers. Moreover, it can yield ultrafast chaotic dynamic without feedback or subject to short feedback delay, thus forming very short spacing between two virtual nodes under sufficient nodes, denoting two reservoir computers using two chaotic polarization components of optically-pumped spin-vertical-cavity surface-emitting laser can deal with two high-speed chaotic time-series in parallel data. In this work, we utilize two parallel reservoir computers using the two polarization components of an optically pumped spin-vertical-cavity surface-emitting laser with both optical feedback and optical injection, to model the chaotic dynamics of the output two polarization components from another optically pumped spin-vertical-cavity surface-emitting lasers as a transmitter. High-quality chaotic synchronization between a transmitting polarization component and its corresponding trained reservoir can be realized by training vertical-cavity surface-emitting laser-based reservoir. Under such a synchronization condition, we demonstrate the successful dual-channel secure communications with 16QAM messages under guaranteeing their securities. We further discuss the bit error ratio performances for two decoded messages under different parameters. We demonstrate that all bit error ratio via different parameters keep at 0. Our findings show that a delay-based optical reservoir computing provides an effective method for the practical application of optical secure communication.
Fiber lasers have a wide range of applications in material processing, biomedical, industrial production, military security and other fields with high electrical-optical and optical-optical conversion efficiency, variable wavelength range, high quality of output beam, compact structure, and low cost, which becomes the research hotspot in recent years. Fiber lasers with the ultrafast laser pulse are mainly realized by active mode-locking and passive mode-locking methods. Compared with the active mode-locking technology, the passive mode-locking has attracted more attention on account of its advantages such as narrower pulse output, compact structure and convenient operation, in which saturable absorber is the core component. The modulation depth and saturation fluence of saturable absorbers determines the output characteristics of passively mode-locked lasers. Therefore, it is necessary to explore excellent saturable absorbent materials. Ge2Sb1.5Bi0.5Te5 materials have a wide spectral response, high thermal, chemical and mechanical stability, and can achieve rapid transition between amorphous and crystalline states at low temperature. Ge2Sb1.5Bi0.5Te5 materials have been frequently used in laser direct writing technology as an outstanding photoresist, but its research in the field of photonics is still in its infancy. To our best knowledge, Ge2Sb1.5Bi0.5Te5 materials have not been used as the saturable absorber for the generation of the ultrafast laser pulse. Therefore, the exploration of optical properties and applications of Ge2Sb1.5Bi0.5Te5 could promote the comprehensive understanding of Ge2Sb1.5Bi0.5Te5 materials and effectively drive the development of optical devices based on Ge2Sb1.5Bi0.5Te5 materials. Here, Ge2Sb1.5Bi0.5Te5 films with the thickness of 40 nm were prepared by the magnetron sputtering method on a gold mirror under the atomspheric pressure of 0.1 Pa, the power of 50 W and the Ar flow of 25. Then the Ge2Sb1.5Bi0.5Te5 films were annealed at 150 ℃ for 20 min in vacuum tube annealing furnace. X-ray diffraction analysis of Ge2Sb1.5Bi0.5Te5 films shows that the as-grown Ge2Sb1.5Bi0.5Te5 films are amorphous and after annealing the crystalline state has been observed in Ge2Sb1.5Bi0.5Te5 films. The optical absorption of crystalline Ge2Sb1.5Bi0.5Te5 films in the range of 1 300~1 600 nm is significantly higher than that of amorphous Ge2Sb1.5Bi0.5Te5 films, and the absorbance of amorphous and crystalline Ge2Sb1.5Bi0.5Te5 films at 1 550 nm are 5.61% and 21.16% measured by spectrophotometer, respectively. A typical balanced twin-detector test system is used to explore the nonlinear absorption characteristics of amorphous and crystalline Ge2Sb1.5Bi0.5Te5 saturable absorber. The modulation depth of the amorphous Ge2Sb1.5Bi0.5Te5 saturable absorber is 2.7% and that of crystalline Ge2Sb1.5Bi0.5Te5 after annealing increases to 3.8%, which is 1.4 times higher than the amorphous Ge2Sb1.5Bi0.5Te5 saturable absorber. The reason for the improved nonlinear absorption capacity of crystalline Ge2Sb1.5Bi0.5Te5 films can be mainly attributed to the formation of the long-range ordered structure with less unsaturated bonds and less local defects in the crystalline Ge2Sb1.5Bi0.5Te5 saturable absorber. Moreover, the crystalline Ge2Sb1.5Bi0.5Te5 saturable absorber also exhibits the smaller band gap which decreases the binding energy of excitons. The lower binding energy of excitons slows the recombination efficiency of electron hole pairs, and the modulation depth of saturable absorber is inversely proportional to the recombination efficiency of electron hole pairs. Therefore, crystalline Ge2Sb1.5Bi0.5Te5 with a smaller band gap has a greater modulation depth. Based on the crystalline Ge2Sb1.5Bi0.5Te5 saturable absorber, an Er-doped fiber laser system has been constructed to obtain the ultrafast laser pulse in which a semiconductor laser with a central wavelength of 980 nm is used as the pump source and is connected to a wavelength division multiplexer at 980/1 550 nm. A 1.3 m Erbium-doped fiber is employed as the gain medium. The polarization independent isolator ensures the unidirectional operation of the signal light in the cavity, and the polarization controller is used to adjust the polarization state in the cavity to affect the mode-locked state. The laser beam is applied to saturable absorber through the ring device and re-coupled back to the ring cavity to generate a mode-locked laser pulse. Then the laser pulse passes through a fiber coupler with a beam splitter ratio of 90∶10. 10%of the output laser is used to observe the performance of the laser in real time, and the remaining 90% of the laser is used to continue oscillating in the cavity. The obtained output pulse width is 1.52 ps, the repetition frequency is 3.44 MHz, and the signal-to-noise ratio is 47 dB. In addition, to investigate the effect of film thickness on the absorption properties, Ge2Sb1.5Bi0.5Te5 films of 60 and 80 nm were prepared by magnetron sputtering. The experimental results of the UV-Vis-NIR absorption spectra show that the optical absorption increases with the increase of Ge2Sb1.5Bi0.5Te5 films thickness, which indicates the controllability of optical properties of Ge2Sb1.5Bi0.5Te5 films and reveals the great potential of Ge2Sb1.5Bi0.5Te5 materials in ultrafast lasers. Our finding can provide reference for the application of phase transition materials in photonics and help for the wide application of fiber lasers.
Semiconductor laser is one of the very important sources in a variety of modern physics experiments, the measured quantities in quantum precision measurement will be demonstrated to the laser parameters such as amplitude, frequency, and phase, so researching semiconductor laser’s phase noise and amplitude noise is significant in the domain of phase-sensitive amplification and interferometric phase measurement. As an optical cavity to improve beam quality and beam mode, mode-cleaner cavity has important applications in quantum precision measurement, quantum communication, quantum computing, quantum key distribution and gravitational wave detection.The noise conversion functions of phase and amplitude fluctuation of a light beam can be deduced, according to the properties of reflection field and transmission field for the mode-cleaner cavity. The mode-cleaner cavity and the balanced homodyne detection are constructed after locking the laser’s frequency. The balanced homodyne detection measures amplitude noise of reflection field and transmission field, the experimental results show that the amplitude noise of transmission field decreases gradually with the detection frequency, and the noise at the frequency far larger than the linewidth reaches the shot noise limit. Nevertheless, the amplitude noise of the reflection field is far larger than that of the transmission field. As a whole, the amplitude noise also decreases gradually with the detection frequency. Linewidth of mode-cleaner cavity is calculated as 2.5 MHz by transmission spectrum, mode cleaner can be regarded as a low-pass filter under the condition of meeting the resonance in the transmission field, the high-frequency noise is greatly suppressed. In this case, when the detection frequencies range from 0 to 2.5 MHz, the amplitude noise of the transmission field is far larger than the shot noise limit. In the range of detection frequencies from 2.5~20 MHz, the amplitude noise decreases gradually with the increase of the analysis frequency in the transmission field. When the detection frequency is 15 MHz, the amplitude noise of the transmission field reaches the shot noise limit. However, when the detection frequencies range from 0 to 20 MHz, the amplitude noise of the reflection field is far larger than the shot noise limit. On the one hand, the mode-cleaner cavity can be regarded as a high-pass filter in the reflection field, which makes a significant contribution to the results. On the other hand, mode-cleaner cavity can interconvert the phase noise and amplitude noise after reflection or transmission. The amplitude noise of transmission field and reflection field can be calculated theoretically based on the noise conversion functions. By comparing the theoretical results with experimental results, the experimental results are consistent with the theoretical results.The phase noise of semiconductor laser is analyzed by measuring the amplitude noise of the reflection field with on-resonance locking and near-resonance locking. The results indicate that the experimental results are consistent with the theoretical results under the condition of on-resonance locking and near-resonance locking. Thus the research demonstrates that phase noise of semiconductor laser can be measured and evaluated by the noise conversion function. Moreover, this study provides a theoretical basis and provides an important complement for the noise conversion of mode-cleaner cavity. The mode-cleaner cavity is studied from the point of noise conversion in this paper, and the research demonstrates that the experimental results are consistent with the theoretical results in the transmission and reflection field. At the same time, the phase noise of semiconductor laser is measured and evaluated, which promotes the development of quantum precision measurement.
With the vigorous development of aerospace technology, space communications, especially microwave optical transmission systems, require high-power, high-efficiency lasers and amplifiers to meet the requirements of high speed, large capacity and long transmission distance in space communications. The inter-satellite microwave photonics subsystem transmits the microwave signal to another satellite by modulating the microwave signal to the laser. The signal loss is very large after long-distance transmission. It is necessary to use a fiber amplifier as a preamplifier to improve the signal-to-noise ratio. The conversion efficiency of the fiber amplifier is a key performance index, which is related to the power efficiency, heat dissipation and weight of the entire laser amplifier. Aiming at the space application of optical fiber amplifiers, the design and experimental research of optical fiber amplifiers with compact structure, high output power and low power consumption have been carried out. In the electrical design, a semiconductor laser drive system with a digital signal processing chip as the core processor is designed, and a closed-loop control logic of output-detection-re-output is formed. Through automatic power control, the stability of the output power of the pump laser is ensured, the power consumption of the pump source during idle time is reduced, and the electro-optical efficiency is improved. The drive system is equipped with a temperature sensor and a semiconductor cooler to monitor temperature data in real time, and the semiconductor cooler will maintain the temperature stability of the pump laser. The powers of the input signal light and the output signal light can be obtained by two photodetectors, and the real-time sampling and comparison algorithm realize the automatic standby system when there is no input, and maintain the minimum power consumption operation. The mechanical design adopts photoelectric separation and modular design. Taking into account the temperature-sensitive characteristics of semiconductor lasers and the particularity of heat dissipation in the space, the optical part and the electronic part are respectively integrated in different spaces. The heat pipe in the heat dissipation layer will transfer the heat generated to the external heat treatment equipment, so that all devices do not need to consider too much heat dissipation space or add the heat sink, keeping the overall structure compact and small, and maintaining good heat distribution. The integrated design reduces fixed parts and keeps the overall weight below 1 kg. The optical structure adopts a two-stage amplification structure design, and the signal light is amplified through the erbium-doped fiber and erbium-ytterbium co-doped fiber, so that high power output can be maintained while maintaining a high signal-to-noise ratio. Furthermore, this paper studies the output performance of the entire system under different pump powers, and compares the conversion efficiency, power consumption and spectral characteristics of the amplifier under different fiber lengths. In order to obtain a good gain and signal-to-noise ratio, an 18 m long erbium-doped fiber and a 7.5 m erbium-ytterbium co-doped fiber are selected, the output spectrum of the signal light after the two stages are analyzed. Finally, the 1 551 nm milliwatt seed light is amplified to a 7.39 W signal output, and the signal-to-noise ratio is maintained at 40 dB. The output beam quality is 1.05, and the overall power consumption does not exceed 80 W. In this work, all the optical designs use polarization-maintaining devices, and the polarization extinction ratio of the fiber amplifier is 17.5 dB. The computer can be used to change the injection current of the pump source, thereby changing the final output power. All the modules can be specially adjusted according to the environment, and can be flexibly used in various occasions. This compact fiber amplifier is 20 cm long, 14 cm wide and 4 cm high, which is very suitable for space laser communication.
The air motion parameters (such as air speed, angle of attack, angle of sideslip, etc.) of an aircraft, which are important source parameters for flight control, navigation, and mission decision making, are usually measured by traditional airborne air data system, and can be used in flight stability control, accurate navigation and precise weapon launch. But the traditional air data system can no longer meet the performance requirements of modern military and civil aircraft in maneuverability, stealth, reliability, safety, comfort and economy due to the performance defects, such as measurement failure at low air speed and large maneuvers, significant aerodynamic delay, pitot tube icing, poor stealth performance, susceptibility to aircraft air turbulence, requiring complex compensation, etc., which result from its mechanical, nonlinear, near the fuselage measurement characteristics. To overcome the defects of the traditional air data system, the method using laser technology for measuring air motion parameters was first developed abroad, which can completely solve the defects of traditional airborne air data system due to its characteristics of high accuracy, high linearity, measurement far away form fuselage, embedded installation, etc..In this paper, the principle, characteristics, advantages and disadvantages of the traditional method and laser method for measuring air motion parameters are compared and analyzed. The obvious advantages and advancements of the laser measurement technology for air motion parameters make it a research hotspot at home and abroad. Throughout the technology development, the laser measurement technology for air motion parameters can be categorized two schemes including direct detection and coherent detection. The principle and composition of the two technical schemes are presented. Then the advantages and disadvantages of the two technical schemes are compared.The two technical schemes of laser measurement technology for air motion parameters, which have their advantages respectively, are developed synchronously nowadays and have been widely used. The application areas mainly include three aspects: accurate measurement of air motion parameters, flight calibration of conventional air data system and detection of wind shear and turbulence ahead of the aircraft. In this article, the development and application are mainly reviewed in these three aspects, and the achievements of prototype and flight test results in these three application aspects are provided. As can be seen from the research reports, the research institutes are mainly concentrated in Europe and America, including OADS corporation, Ophir corporation, Michigan Aerospace corporation, Thales corporation, ONERA, EADS, DLR, etc. Many flight tests have been carried out and a large number of test data have been accumulated by these institutes. At present, the prototype which can be equipped and applied has been successfully developed abroad. In contrast, domestic research is relatively backward, the main research institutions are AVIC CAIC and AVIC FACRI. In recent years, the principle prototype has been reported by the two research institutes respectively.The development direction of the two schemes of air motion parameters measurement technology are prospected respectively. The coherent detection scheme is likely to be first equipped for airborne application due to its low size, weight and power property. The detection capability of the coherent detection scheme at high altitude up to tens of kilometers needs to be continuously improved. The direct detection scheme should further reduce the size, weight, power and costs to meet the requirements of airborne applications. The application of quantum technology in laser measurement technology for air motion parameters is prospected. In the future, single photon detection technology and quantum enhancement technology based on compressed state photon are expected to be applied to improve the detective sensitivity of the system and achieve ultra-sensitive detection. In view of the huge development gap between home and abroad, some useful suggestions are provided for domestic research of laser measurement technology for air motion parameters, such as robust related industrial chain, focus on low SWaP performance design, strengthen cooperation, increase capital investment, etc.The purpose of review and further clarification of the principle, application and development trend of laser measurement technology for air motion parameters is to provide a useful reference and new ideas for the researchers engaged in prototype and application research, and to promote the in-depth application of laser measurement technology for air motion parameters in aviation fields.
As the key equipment of the starlight navigation system, the starlight direction finder is mainly used to obtain the accurate spatial angular position of natural stars, which is used for high-precision attitude calculation and navigation and positioning of missiles, ships and space satellites. Its performance largely determines the performance of the starlight navigation system. In the visible light band, the starlight direction finder is easily affected by the background radiation generated by the sun scattering in the atmosphere. According to Rayleigh scattering theory, the infrared band has a smaller scattering intensity than the visible light band, and the radiation energy transmitted through the atmosphere is strong. Practice shows that it is much more efficient to use the near-infrared band to measure stars in daytime than the visible light band to realize navigation and positioning. Therefore, developing high-precision infrared star simulator and realizing infrared star map simulation has become a new content to improve the efficiency of starlight navigation and the accuracy of star point detection. In order to realize the ground calibration of starlight director and meet the three simulation requirements of radiation uniformity, specific irradiance and spectrum, this paper proposes an optical system design scheme of infrared star simulator with uniform radiation under specific irradiance. According to the requirements of irradiance simulation and uniform irradiation, the radiation flux transfer model of light source system is established, and a uniform radiation light source system is designed to meet the requirements of radiation spectrum. According to the docking requirements of starlight direction finder and infrared star simulator, a transmission collimating optical system with high imaging quality is designed. Combined with the design results, Tracepro software is used to model the optical system, and the irradiance and uniformity of the radiation surface of the light source system and the exit of the optical system are simulated and analyzed. The irradiance and uniformity of the radiation surface of the light source are tested experimentally with an irradiance meter to verify the accuracy of the theoretical analysis. The measured results show that the irradiance of the radiation surface of the light source meets the index requirements, and the irradiance unevenness is 2.75%, which meets the development requirements of the infrared star simulator.
This paper describes the design of the entrance window assembly for the Full-disk Magnetograph (FMG) which is one of the three main payloads of the Advanced Space_based Sloar Observatory mission (ASO-S). The entrance window plays the role to reduce the effect of external space environment on the After Optical System(AOS), to transmit the visible light, and to prevent the infrared radiation and pollution, hence, it is one of the most important components in the design of space solar observatory payloads, as the entrance window is facing the sun all the time on orbit, both its design and verification test are very challenging task.A heat-rejecting entrance window assembly of FMG is designed to ensure that the imaging quality of the whole system is not affected by the drastic change of space environment and the temperature distributions can meet the thermal requirements, then the mechanical, optical and thermal design processes are briefly described. FMG is sun-oriented on orbit with long-term operation, and its optical system is transmissive. Thus, its mirror temperature will be more sensitive to the given solar radiation parameters than the other space optical payloads, and it is necessary to correct solar radiation parameters to improve the accuracy of simulation. Thermal analysis method with equivalent solar radiation parameters is studied to evaluate the heat-rejecting ability of the HEWA, and three analysis cases are selected based on the heat flux of the HEWA and the operational mode of FMG on orbit, while, Case1 and Case 2 are normal operational modes, and Case 3 is a calibration mode.Only 5 nm wide transmission pass-band around the science wavelength (532 nm) is able to reach the AOS of FMG because of the spectral selectivity of the window glasses. The thermal balance test of the HEWA not only needs to simulate the solar radiation intensity, but also needs to simulate solar collimation and spectral characteristics accurately. The solar simulator can simulate the solar radiation intensity, collimation and spectral characteristics adequately, it has higher heat flux simulation accuracy than other methods. Thus thermal balance test with solar simulator is carried out to verify the design and the analysis of the HEWA. Three test cases which are consistent with the analysis cases are carried out during the thermal balance test, quantitatively speaking the analysis results coincide with the test results, however, some main differences exist between them: 1) Temperatures of the baffle during the test are always lower than that of the analysis; 2) The test temperature of the window glasses in normal operational modes are always higher than that of analysis; 3) The temperature difference of Mirror1 in Case 3 is larger than that of the analysis.Some modifications made on the analysis model make the numerical analysis results being quantitatively consistent with the test results. Then the modified analysis model is used to predict the actual on-orbit temperature distribution of the HEWA. From the numerical results, it is found that only about 0.134 W solar radiation is able to pass through the HEWA and be absorbed by the primary mirror of the AOS, the maximum temperature of the window frame on-orbit is 28.2℃, and that of the window glasses is 26.3℃, while the primary mirror of the AOS is able to maintain at 22±2℃. Thus the designed HEWA of FMG is able to withstand each typical condition on orbit and meet the requirements of the mission. It avoids that the optical performance of the space solar observatory payload will decrease or the optical system will be polluted because of the overhigh temperature of the entrance window assembly and the AOS on orbit, which is able to guide the design of transmission optical system and other optical payloads for solar observation.
The attitude measurement data in the vertical take-off phase of a rocket is of great significance to analyze the running orbit, aerodynamic parameters and flight control performance of rocket. The traditional attitude measurement of rocket vertical take-off phase mainly includes telemetry, optical measurement, and radar measurement. The violent vibration has a great impact on the attitude measurement accuracy of telemetry, and once the rocket takes-off fails, it is difficult for telemetry method to obtain effective original analysis data. Although the optical measurement accuracy is high, it needs to use multi station optical equipment to interpret the rocket attitude data after rendezvous, so the real-time performance is poor, and the optical equipment is vulnerable to the interference of weather environment and tail flame during take-off phase, so there is a risk of missing rendezvous data. Although radar measurement is little affected by weather conditions, it is easy to be disturbed by ground clutter. Therefore, there is no external real-time attitude measurement data in the rocket vertical take-off phase. It is urgent to fill the data gap in this phase through new measurement methods to ensure the safety of the rocket vertical take-off phase.Aiming at the technical problems of external real-time attitude measurement in rocket vertical take-off phase, considering the advantages of Lidar, such as high precision, all-time measurement, high resolution and not easily disturbed by environment, the real-time attitude measurement method in rocket vertical take-off phase based on Lidar is proposed in this paper. Raytheon intelligent MS03-A500 four wire Lidar is adopted, and the Lidar is installed on the two-axis tracking frame to form the measurement system. Before the rocket is launched, the Lidar continues to scan the middle and upper part of the rocket to obtain the static laser point cloud data, correct the point cloud data and solve the spatial coordinates, and adopt the multi ellipse center fitting central axis algorithm. It is calculated and analyzed that the static and dynamic attitude measurement accuracy of Lidar are 0.018 8° and 0.049 8° respectively. In the rocket launch test, the Lidar measurement system is arranged at a launch site 150 meters away from the rocket. In the rocket vertical take-off phase, the Lidar tracks and scans the fixed position of the rocket with high precision, and obtains the rocket attitude change value with real-time and high precision.To verify the reliability and rationality of Lidar measurement data, three sets of optical equipment are used to measure the rocket attitude angle intersection at the same time. After comparing the rocket attitude change values measured by Lidar and optical equipment, the following conclusions are drawn: 1) According to the measurement accuracy calculation results and the verification of test data, the attitude measurement accuracy based on Lidar is about 5 times higher than that of optical measurement equipment. 2) Considering the different measurement accuracy of the two equipment, the variation trend of rocket yaw angle and pitch angle measured by Lidar and optical equipment is basically the same in this test, which verifies the correctness and rationality of the attitude measurement methods and measurement accuracy of the two kinds of equipment. 3) The real-time high-precision attitude measurement and data output of the rocket based on Lidar is realized, which effectively fills the gap of a real-time attitude measurement outside the rocket, provides a real-time attitude data source for security control equipment, and ensures the launch safety of the rocket.
Digital Speckle Pattern Interferometry (DSPI) provides an effective means of full-field and non-contact measurement of deformation or displacement. With the advancement of the aerospace and automotive industry, deformation measurements with a large Field Of View (FOV), high resolution, and wide measurement range are becoming more and more urgent. However, it is difficult to increase the FOV for a given size of CCD without compromising the lateral resolution of the deformation measurement. To solve this problem, a technique for stitching the phases of multiple sub-images to enlarge the FOV without impairing the lateral resolution was investigated. The existing aperture synthesis methods usually obtained multi-images by moving CCD or object. They are only applicable to the measurement and observation of stationary objects. For deformation measurements, at least two surface states of the object are involved, corresponding to before and after deformation. Thus, the positioning errors and axial misalignment between corresponding hologram pairs are difficult to estimate. To overcome the disadvantages of multi-step image acquisition schemes. An experimental setup with multiple CCDs was constructed to obtain multiple sub-images. The phase of each CCD was extracted by the Fourier-transform method, and then the unwrapped phase maps of the overlapping areas were used to estimate the relative positions. Subsequently, the phase deviations between adjacent sub-image pairs were estimated and compensated for correct phase stitching. In order to obtain the largest possible FOV using as few CCDs as possible, the effect of the size of the overlap area on the stitching results was analyzed. The relationship between the standard deviation and the size of the overlapping area was investigated. The standard deviation is less than 0.015 μm when the size of the overlap area is between 141 and 461 pixels, corresponding to a percentage of the overlap area between 8.8% and 28.8%. Therefore, the size of the overlap area is approximately 10%, which may be appropriate in terms of the trade-off between FOV and accuracy. With the proposed method, the FOV was expanded from 5.5 cm×4 cm to 10 cm×4 cm and only two CCDs were used. The maximum relative error before and after stitching of the overlapping area was less than 1%, which illustrates the effectiveness of the proposed method. In addition, to further demonstrate the effectiveness of the phase stitching method, a calibrated loading device (the loading range is 0~9 μm , the expanded measurement uncertainty is 0.2 μm with the coverage factor k=2) is driven by a piezoelectric actuator was used. A total of 9 displacement loading points were included, and three groups of values were measured by CCD#1, CCD#2, and the phase stitching method. The Least-Square (LS) method was used to fit the measured deformation of the three groups and the fitting residuals were evaluated. Additionally, the coefficient of determination R and the Root Mean Square Error (RMSE) of the quality of the fitting were compared. The measurement accuracy of the phase stitching method was equivalent to that of the single-camera method when comparing the measurements of the calibration points by the Root Mean Square Error (RMSE) metric. In summary, the proposed phase stitching method based on multi-CCDs deformation measurement is an effective means to increase the FOV without impairing the lateral resolution. At the same time, with a certain FOV, the measurement range and axial resolution can increase. Theoretically, for the deformation distribution similar to the cantilever beam, the measurement range can increase with the increment of FOV.
With the development of lithography technology in China from deep UV to extreme UV, the detection accuracy of lithography objective lens surface shape below nano or even sub nano is required. The detection accuracy of the commercial Tyman-Green and Fizeau interferometers is limited by their own standard reference mirror shape accuracy, which has been unable to meet the development needs of the above-mentioned ultra-precision detection. The point diffraction interferometry uses the micro aperture diffraction to produce an almost ideal spherical wave as the reference wave, which gets rid of the limitation of the measurement accuracy of the standard reference mirror, and opens up an effective way for the ultra-fine surface shape detection in the field of lithography. The point diffraction method was proposed in the early 19th century and developed rapidly abroad, and the United States and Japan have developed molding equipment for point diffraction interferometers. The research on point diffraction interferometer in China started late, especially in the field of small hole point diffraction interferometry. At present, the research is mostly in the laboratory stage. One of the most important reasons restricting the improvement of its accuracy and the development of forming equipment is its phase shift. The existing small hole point diffraction interferometers mostly use high-precision piezoelectric ceramic phase shifter to drive the measured part to move for many times, and collect multiple interference images for phase extraction. The realization of this kind of time-domain multi-step phase-shifting technology relies on the import of expensive PZT on the one hand, and on the other hand, the interference image is easily affected by environmental interference during the long-term acquisition of interference images, resulting in low actual detection accuracy, especially low repeatability accuracy. In order to solve the problem that the phase shift depends on the piezoelectric ceramic phase shifter in the small hole point diffraction interferometer and the measurement is affected, a super precision surface transient interference detection method based on small hole point diffraction is proposed, and the system feasibility is verified from two aspects of theoretical analysis and comparative experiment. Two orthogonal circularly polarized beams are obtained by constructing a small hole diffraction interferometry optical path with reflective structure. The checkerboard phase grating is used to split the light and the polarizer array is phase-shifted, and four interference images with different phase shifts are simultaneously obtained on the CCD. The shape information of the measured surface can be obtained directly by processing the four instantaneous interference images collected at a single time. The theoretical analysis results of the system polarization state based on Jones matrix theory show that when the polarization angles of the polarization array are 0°, 45°, 90°and 135°, the introduced phase shifts are 0, π/2, π and 3π/2 respectively, which meet the requirements of the four-step phase extraction algorithm. The system can realize the transient detection by splitting the light through the chessboard phase grating and phase shifting with the polarizer array; Compared with the Zygo interferometer, the experimental results show that the measurement results of the proposed system are close to those of the Zygo interferometer, and the repeatability and accuracy of PV are better than those of the Zygo interferometer λ/200, the repeatability accuracy of RMS is better than λ/1 600.The experimental results show that the system is feasible and has good stability. The proposed system takes into account the advantages of high precision of point diffraction and strong anti-interference ability of synchronous phase shift. The principle is simple and easy to implement. The research results are conducive to promoting the process of small hole point diffraction interferometry from experimental research to instrument development.
Micro-computed tomography is widely used for medical diagnosis and non-destructive testing since the 1970s. In three-dimensional micro-CT imaging, the quality of the reconstructed images depends highly on the precision of the geometric parameters of the system. Image quality could be severely degraded by misaligned parameters, resulting in artifacts, blur and lower resolution in images, especially for CCD-coupled CT system (CCT system). Therefore, high-precision methods for estimating CT system parameters are necessary. Different research groups have proposed various methods to estimate geometric parameters for fan-beam CT or cone-beam CT. In terms of data acquisition methods, the main methods can be categorized as multi-angles projection and single-angle projection. However, due to the narrow field of view caused by the ultra-high resolution of the CCD-coupled detector, neither of these two methods can be applied to CCT system directly. Generally, a CCT system consists of an X-ray source, a CCD-coupled detector, and a precision rotation stage for placing samples, in which the X-ray source and detector are distributed on both sides of the sample. During scanning, the sample rotates around the rotation axis, while the X-ray source and the detector remain stationary. In an ideal CCT system, strict geometric alignment must be satisfied. It can be summarized as follows: Firstly, the X-ray focal spot, the axis of the rotation, and the center of the detector should be in a plane; secondly, the straight line which passes the focal spot and the detector center, should be perpendicular to the axis of rotation and the detector; thirdly, the axis of the rotation should be parallel to the row of the detector. Basing on precious studies, there are five geometric parameters to be calibrated totally in a CCT system. We conclude all the calibrated parameters as follows: 1) u0,v0 are projection coordinate of the X-ray source spot on the detector; 2) η is the angle of detector tilt around X axis; 3) R is the distance between X-ray source and rotation axis; 4) D is the distance from X-ray source to detector.In this paper, a mathematical model based on a double micro-sphere phantom is proposed to estimate all the above geometric parameters of CCT system. The double micro-sphere phantom consists of two spheres with a diameter of 100μm. Generally speaking, when a sphere rotates around the rotation axis, the centers of its projections on the detector form an ellipse or a line. To estimate all the parameters, the calibration is carried out in two steps. Firstly, the circular scanning imaging of the double micro-sphere phantom is carried out to obtain its elliptical projection trajectory. Then the mathematical relationship between the ellipse equation corresponding to the trajectory and the geometric pose parameters of CCT system is established to solve the geometric pose parameters.To verify the effect of the calibration method in realistic imaging system, an experiment was performed in a misaligned CCT system. After the parameters calibrated using the mathematical model proposed in this paper, a bamboo fiber was scanned by the CCT system. Then, the results of the calibration parameters were substituted into the reconstruction algorithm, and the reconstructed images were compared before and after parameters calibration. Experimental results show that by substituting the corrected geometric parameters solved by the mathematical model into the reconstruction algorithm, the aliasing and blurring of edge information in the reconstructed images are reduced yielding an improved image detail resolution. In addition, we adopted the Energy of Gradient (EOG) to quantitatively evaluate the image quality. The EOG value of reconstructed image after calibration is larger than before, which also shows the image quality after calibration is improved.This paper proposes a new method to estimate geometric parameters of CCT system. The method allows us to estimate all the parameters by acquiring projections of the designed double micro-spheres phantom. Compared with other methods, it could be used to CCT system which has high resolution and narrow field of view. Meanwhile, this method removes the influence of rotation errors on geometric pose parameters. The accuracy of this method is confirmed by experimental results. At last, a bamboo fiber sample was scanned and reconstructed. The experimental results show that the artifacts are reduced and image quality is improved.
The production process of Micro-camera module includes plate washing, plate drying, glue painting, crystal pasting, bonding, glue baking, detection, etc. The glue painting is an important step and its quality has an important effect on the mounting performance of the Micro-camera module. Three-dimensional measurement is the key technology to ensure its mounting quality, and usually, the quick glue painting process needs the accuracy and agile detected method. To meet the real-time measurement requirements of glue surface in the mounting process of Micro-camera module, the height information mapping model of the detected glue surface is established based on monocular vision imaging system, and a three-dimensional measurement method of glue surface is proposed in this manuscript. The proposed method can be divided into following three steps. Firstly, by analyzing the influencing factors of visible light focusing imaging system and describing the image features based on the mentioned monocular vision system, the imaging model is built, which reveals the influence rule of light source intensity, object surface inclination, diffuse reflection coefficient and reflected light deflection angle on image gray value in such imaging model. Secondly, the mapping model of the surface height information of the detected object is established, and the height calculating equations of the visual surface are derived based on the proposed mapping model. Especially, the proposed equations only contain two un-knowns variables, and the three-dimensional information of the detected surface is obtained by solving the equations.Furthermore, because of the proposed one-to-one mapping model, the theoretical analysis results show that the proposed method can avoid the solution uncertainty problem arising from Shape-from-shading method (SFS method) and the matching problem arising from binocular vision method. Finally, the validity of the proposed method was verified by the glue measurement experiment. In this process, three different detected methods are compared with the same testing condition. Experimental results showed that the proposed method can recover the three-dimensional morphology of the glue surface, and its average height measurement error is less than ±10 μm with the single image detection speed of less than 0.2 s, which verifies the effectiveness and real-time performance of the proposed method.
The demand for high-accuracy measurement of micro-angle in modern industries is getting higher and higher, and its measurement methods and measurement technologies are also constantly improving. At present, The micro-angle measuring instrument with the highest accuracy in the world is the ELCOMAT HR photoelectric autocollimator produced by M?LLER-WEDEL in Germany, and its angle measurement uncertainty within the range of 300″ can reach 0.06″ (k=2). And the highest accuracy in China is the AUTOMAT 5 000 photoelectric autocollimator produced by Tianjin Automate Optoelectronics Co., which can achieve a measurement accuracy of ±0.25″ within the range of ±1 000". With the continuous improvement of the precision of micro-angle measurement, higher and higher requirements are put forward for the angle calibration, and some traditional angle calibration methods are difficult to meet the current needs. Therefore, using the self-calibration technology to realize micro-angle measurement has become a research hotspot in recent years. The current angle self-calibration technology mainly focuses on the measurement of the circumference angle with the characteristic of circle closure, and there are few studies on the self-calibration measurement of micro-angle. In the early stage, our research group proposed a micro-angle measurement system based on F-P etalon, which used the displacement of the concentric rings in the focal plane after F-P multi-beam interference ring imaging to achieve micro-angle measurement, and pointed out the possibility of using the system to realize the micro-angle self-calibration measurement. This paper provides a comprehensive review and summary of the self-calibration method for this micro-angle measurement. The key point of this self-calibration method is to use the exact fraction method to measure the exact value of the F-P etalon interval, and then accurately calculate the relative focal length of the imaging objective lens, and combine the relative displacement caused by the small angle, so as to realize the self-calibration of the micro-angle measurement. The main research work of this paper is as follows: 1) The principle and method of the self-calibration measurement of the micro-angle measurement system are systematically and detailedly sorted out, and the calculation method of exact fraction method is described in detail, and the complete micro-angle self-calibration measurement process is finally obtained. 2) Special consideration is given to the effect of the algorithm error of the exact fraction method, temperature and humidity on the measurement results of the F-P etalon. The interference image under the theoretical interval d0 is obtained through simulation by MATLAB, and the calculated interval d1 of the F-P etalon under this condition is obtained by using the exact fraction method, and compared with the theoretical interval d0, the algorithm error and the temperature and humidity error of the exact fraction method are obtained, and the correction of the calculation result of the F-P interval is realized. 3) The self-calibrated micro-angle measurement experiments are carried out, and the accurate interval of the F-P etalon in the current environmental conditions is measured. Combined with the simulation method in 1), the corrected value under the current environmental conditions is obtained, and finally the corrected etalon interval is obtained, which is d '=(2 014.986 5±0.000 3) μm. The focal lengths of the imaging objective and the micro-angle measurement results before and after self-calibration are obtained. The measurement results show that under the current experimental conditions, the relative expanded measurement uncertainty of the focal length after self-calibration has reduced from 0.014 to 0.007, while the angle measurement uncertainty in 600" has decreased from 0.132″ to 0.084″, which promotes the accuracy of micro-angle measurement greatly.
The optimization target of traditional light-weight methods is a single part generally, which can not make the whole component achieve the design goal of lightest weight and highest surface figure precision. It can not meet performance indexes due to the strict limitation of the carrying capacity of actuator and mirror's material. A Primary Mirror Segment Assembly ( PMSA ) is designed integrated to reduce the load of the segment actuator and ensure the surface figure and stability of the splicing primary mirror. The points surpport on the mirror back is selected as the support method through the topology optimization, the mass distribution of the back of the mirror and the distribution of the removable part are obtained. The initial structure of the mirror is determined based on back with triangular hole for light-weighting. A multi-axis flexible support structure with slotted beams based on the support method is put forward. Component integration design introduces too many parameter variables, resulting in a large amount of calculation, long iteration time and difficulty in convergence. A combinatorial optimization algorithm based on multi-island genetic and gradient algorithm to solve the problem of too many variables and difficulty in convergence is designed, and it is applied to optimization design of the segment. The combinatorial optimization algorithm's tactics is that the sensitive area of the design space is roughly located by the multi-island genetic algorithm, and the exact positions are optimized by using the gradient optimization algorithm in the area. The tactics makes full use of the advantages of global optimization algorithm in overall design space traversal and gradient optimization algorithm in local area convergence speed faster. It is established that multiple-object optimization model which combines structure mass, the surface figure error under the influence of gravity along Y-axis ( dectction direction ) and the surface figure error under the effect of 5℃ temperature change as objectives. The optimization model is carried out on the initial structure of the primary mirror segment assembly, and the statics of the optimized structure under the influence of gravity along Y-axis ( dectction direction ) and 5℃ temperature change are analyzed. The Root Mean Square (RMS) of the mirror’s surface figure error is controlled within 5 nm, which is tested at last and is far better than design requirement. The segment weights 1.74 kg, and primary frequency of X, Y, Z is more than 400 Hz. Its surface density is as low as 27.32 kg/m3. The surface figure error is test and the primary mirror segment assembly with RMS is as low as 0.019λ (λ=632.8 nm ), which meets the design requirement better than λ/50. The design method of integrated optimization is reasonable, and it provides a guiding for design of mirror assembly of space camera. In order to evaluate the optimization efficiency of the combinatorial algorithm, it is compared with multi-island genetic algorithm.The multi-island genetic algorithm is used to optimize the primary mirror segment assembly.The combinatorial optimization algorithm iterates 394 times, and the multi-island genetic algorithm iterates 601 times. The result shows that combinatorial optimization algorithm based on multi-island genetic and gradient optimization has better solving ability and can effectively improve the optimization efficiency of integral optimization design of the primary mirror segment assembly.