This study investigates the effect of aerosol on the polarization distribution of the atmosphere. The polarization distribution patterns under different aerosol conditions are simulated and studied based on the vector radiation transmission mode of the Monte Carlo method. An image-based test system is used to test the polarization pattern under an actual atmospheric condition. The effect of aerosol mass concentration on the polarization pattern is analyzed. The results reveal that with the increase of aerosol mass concentration, the degree of polarization decreases, indicating that the aerosol mass concentration considerably affects the degree of polarization. When the mass concentration of PM10 increases from 27 μg/m 3 to 53 μg/m 3 and then to 103 μg/m 3 and that of PM2.5 increases from 9 μg/m 3 to 28 μg/m 3 and then to 66 μg/m 3, the maximum degree of polarization decreases by an average of 25%. In contrast, aerosol slightly affects the angle of polarization, and thus the distribution of polarization angle is stable.

Turbidity is an important indicator for monitoring water environment and water quality, and the satellite remote sensing technology has the advantages of macroscopic spatial coverage and repeated sampling, which is an effective way of monitoring water turbidity. Based on the remote sensing reflectivity of NPP-VIIRS satellite, a water turbidity remote sensing inversion algorithm is developed and applied to the VIIRS satellite data to obtain a long-time series of satellite-derived water turbidity in the Bohai and Yellow seas. The results indicate that the proposed algorithm has a high accuracy with the R2 of 0.97, the root mean square error of 16 NTU, the mean absolute deviation of 23 NTU, and the mean relative error of 34.63%. On a spatial scale, the turbidity distributions are generally high in the near-shore areas and low in the offshore areas. In contrast in the time scale, the water turbidity is at a high level in winter, but the regions with high turbidity shrink in spring. The turbidity is generally at the lowest level in summer and only the coastal waters show high turbidity values. In autumn, the turbidity gradually increases.

This study proposed a two-dimensional wind field inversion method based on the calculation of the correlation among three laser beams. Three beams were emitted in the same plane, and the two-dimensional wind field data were recovered by calculating the correlation among the three echo signals over time. The principle behind the method and the calculation process were described in detail herein. Then, this approach was used to construct a correlation-based two-dimensional wind lidar system, which could achieve the fast two-dimensional wind detection with a spatial resolution of 30 m and a time resolution of 1 s. The influence of lidar angle on the measurement results and the variations in the correlation coefficient curves under different lengths of time sequences were analyzed, and the suitable lidar angle and length of a time sequence were determined. The horizontal wind measurement experiment was conducted to compare the data by the proposed system with the sensing data of wind speed and direction by a wind tower in the detection path. The results show that the data from the proposed system are well consistent with those from the sensor.

Vortex beam arrays have gradually gained attention because of their potential applications in high-power laser synthesis, space optical communication, and laser engineering. Based on a numerical method, namely, the multiple phase screen method, we study the propagation characteristics of a vortex beam array in von Karman turbulence and analyze the light field evolution. The results indicate that the light beam is fused after transmission over a certain distance, and the topological charge of the fused beam is the same as that of the sub-beam in the initial light field. Considering that the expansion of the spot will lead to a loss of topological charge, thereby reducing the communication efficiency, we analyze the influences of various turbulence and beam parameters on the relative beam radius. Among these parameters, turbulence intensity has a great influence on relative beam width, and the other parameters such as topological charge have different degrees of impact. The relative beam radius of the vortex beam array is smaller than that of the single vortex beam.

In this study, the vortex-shaped photoelectron momentum distributions of a hydrogen atom ionized by two time-delayed circularly polarized laser fields are numerically simulated based on the strong field approximation (SFA) theory. Under the action of two time-delayed laser pulses, the electron absorbs photons to overcome the ionization threshold and undergoes photoionization via two different transition channels to reach the continuum states for electron wave-packet interference. The simulation results show that the orientation of the photoelectron momentum vortices is observed to be related with the polarization directions of the two pulses, whereas the number of vortex arms is related with the carrier frequencies. Furthermore, the dynamic Stark effect is a ubiquitous strong field process, which would result in the distortion of the vortex-shaped momentum distributions when considered. On this basis, the clockwise momentum vortices and their distortion are specifically investigated, revealing that the distortion is attributed to the nonlinear properties of the phase associated with the dynamic Stark effect.

A reflection spectrum model of a compound fiber-optic Fabry-Perot (FP) pressure sensor with short length is established, a two-parameter ellipse-fitting cavity length demodulation algorithm is proposed, and the demodulation for FP-cavity lengths in the range of 26-30 μm is simulated. The results show that the maximum error for the demodulated two-parameter ellipse-fitting cavity length is only 0.05 μm. A demodulation system for the fiber-optic FP sensor is established, and the experiments under different pressures are conducted. A 20-kHz demodulation rate is achieved. Thus, the feasibility and real-time performance of the proposed algorithm for demodulating a compound fiber-optic FP pressure sensor with a short cavity are verified.

We study a ring fiber network based multipoint time-frequency dissemination method with high precision. Further, a novel scheme that can realize self-perception for clock difference is proposed; the sensing measurement and compensation for the clock difference are located at remote stations. Time-frequency synchronization can be achieved between the remote stations and the local station. Subsequently, we perform a time-frequency dissemination experiment via a 100-km ring fiber network to study the correction method of asymmetry deviation introduced by dispersion. The multipoint time-frequency dissemination performance is verified under link temperature variation. The measured peak-to-peak value of synchronization error of time-frequency dissemination is smaller than 400 ps, and the root mean square error is smaller than 60 ps. The results show that the remote stations can be accessed anywhere in the ring fiber network.

The fourth-order pulse amplitude modulation signal requires high linearity and a small dispersion tolerance in an optoelectronic device. The bandwidth limitation of the 10G-class optics device severely degrades system performance. This study describes a 50 Gbit/s non-return zero (NRZ) signal transmission scheme using commercial 25G-class optical components. Herein, for a back-to-back and 25-km standard single-mode fiber transmission, the maximum likelihood sequence estimation and decision feedback equalization based on the least mean square algorithm are used for decoding. A comparison of the bit error rate performances under different received optical power conditions proves that the 50 Gbit/s NRZ signal transmission based on the 25G-class optics device can act as a single-wavelength 50 Gbit/s candidate solution.

Spatially modulated full-polarization imaging systems can achieve synchronous detection for full-polarization parameters of targets. The existing demodulation algorithms are only used in monochromatic light detection systems, and they are not suitable for polychromatic light detection systems. The polarization-demodulation algorithm can be modified by finding the position of the central spectrum to obtain the actual carrier frequency. The experimental results demonstrate that the polarization information of the central wavelength can be obtained by utilizing the modified image-demodulation algorithm in broadband spatially modulated full-polarization imaging systems with less than 5% polarization-detection error. The experimental results provide a guild for analyzing the demodulation algorithm of full polarization imaging systems with polychromatic light.

In this paper, a solar compound-parabolic concentrator (CPC), with the simultaneous enhancement of the geometric concentrator ratio and the receiving angle for a circular absorber, is presented. Herein, the surface-shape model of the novel CPC and its mathematical solution are constructed theoretically. The concentration performance of the proposed CPC is analyzed and compared with that of a traditional CPC. The results show that the ordinate value of the surface-shape starting point decreases when the diameter of the circular absorber increases at the same contingence angle of the proposed CPC with a circular absorber. The ordinate value of the surface-shape starting point is -29 mm when the diameter of the circular absorber is 47 mm and the contingence angle is 5.56°. Moreover, the geometric concentration ratio of this novel CPC decreases with increasing the aperture width angle. In contrast, the permissible receiving angle and the average optical efficiency increase with increasing the aperture width angle. When the aperture width angle is 60°, the geometric concentration ratio, the permissible receiving angle, and the average optical efficiency are 1.16, 74.39°, and 86.77%, respectively. Meanwhile, the energy-flow distribution on the absorber surface of the novel CPC is more uniform than that of the traditional CPC.

An image super-resolution network model is proposed based on a multi-scale recursive network herein. The proposed model mainly comprises a plurality of multi-scale feature mapping units, each of which includes a set of feature extraction layers with different scales, a fusion layer, and a mapping layer. The network performs feature extraction directly on an original low-resolution image, which is then reconstructed into a high-resolution image via sub-pixel convolution. In the training phase, the adaptive optimization method is used to accelerate the convergence of the network model. The experimental results show that the proposed algorithm achieves better super-resolution results, significantly improves the subjective visual effects, and sharpens the image texture. The objective evaluation indicators (PSNR and SSIM) of the proposed algorithm on the common test sets such as Set5, Set14, BSD100, and Urban100 are higher than those of the existing mainstream algorithms.

In order to improve the performance of remote sensing scene zero-shot classification by utilizing the complementarity among different image features, a zero-shot scene classification algorithm based on the fusion of image features is proposed, which combines the fusion of image features with zero-shot classification. The analysis dictionary learning is exploited to obtain the sparse coefficients of image features. The obtained sparse coefficients are concatenated as the fused image feature to reduce the redundant information and retain the characteristics of different image features. Supervised information is introduced to improve the discriminability of the model. The fused image feature is structurally aligned with the semantic word vectors to improve the transfer capability for the unseen class scenes. The fusions of image features at the same level and the different levels are carried out on UC-Merced (UCM) and Aerial Image Dataset (AID) remote sensing scenes datasets, respectively. The experimental results show that the overall classification accuracy and time-consuming of our method are superior to those of other zero-shot classification algorithms and general feature fusion algorithms, which proves the effectiveness of our method.

Typically, star images captured in the atmosphere during daylight hours have a strong background and low signal-to-noise ratio (SNR), which makes it difficult for traditional algorithms to extract the star from the images. To improve the recognition rate, we propose an accurate method for simulating star images and train a deep convolutional neural network with a downsampling layer using the simulated images. The trained network can denoise and enhance the star images. Experimental results demonstrate that the proposed method improves the peak SNR by 11.28 dB within an average runtime of 0.2 s, which is significantly less than that of a traditional neural network. In addition, we test the proposed method on the trained network using real star images and find that the improved SNR is 88.9 times greater than that of the existing methods.

To recognize a monitored area automatically for a high-speed railway perimeter-intrusion detecting system, an adaptive image segmentation and recognition algorithm is proposed. The maximum linear feature of each scene is calculated to regulate the adaptive parameters. Moreover, a new combination rule based on the weight of the boundary point and the area size is proposed to rapidly combine the fragmented regions into local areas. A simplified convolutional neural network is designed, the convolutional kernels are pre-trained, and a sparse element is added into the loss function to enhance the diversity of the feature maps. Experimental comparison results indicate that without the graphics processing unit, the pixel accuracy of the proposed algorithm is highest (95.9%), the calculation time is the least (2.5 s), and the number of network parameters is about 0.18×106. The proposed algorithm considers an effective balance among the segmentation precision, recognition accuracy, calculation time, manual workload, and hardware cost of the system. Therefore, the automation and efficiency of the railway perimeter intrusion detection system are enhanced.

The Monte Carlo method is used to simulate the atmospheric polarimetric distribution under water clouds. The results show that the polarimetric response of water clouds in the ultraviolet (UV) band of 360-400 nm is the largest as compared to that of other spectra. Polarization pictures of buildings, clouds, and the sky in the same field of view are taken using UV-visible polarimetric imaging technology. Hough transform is used to divide these pictures, and statistical analysis is applied to each segment. The statistical results show that the relative differences of the polarization degree and polarization angle of the cloud-free and cloudy areas are -14% and 1.6%, respectively, providing the robustness of polarization angle in atmospheric detection. The UV and visible are found to be complementary in polarimetric detection for the clouds. Thus, image fusion technology in conjunction with a Laplacian pyramid can improve the detection capability for atmospheric targets. Results verify that the UV-visible polarimetric imaging technology with large field-of-view and high resolution is feasible and effective for atmospheric detection.

Intensive electromagnetic pulses are common in high-power laser and X-ray equipment and in various environments, such as future battlefields. This kind of harsh environmental factor can affect electronic equipment, especially optical-imaging devices that cannot be shielded well while performing specific imaging tasks. Herein, we choose a digital camera and discrete charge coupled device (CCD) camera system as typical devices used for testing. We investigate the effects on optical-imaging device when the device is subjected to an intensive electromagnetic environment. We observe such effects as equipment malfunctions, decline in imaging quality, and module destruction. Herein, the observed effects and relevant electromagnetic environment data are summarized and analyzed. We promote the idea of using failure-probability thresholds under certain environmental conditions to measure the equipment viability. We provide the probability threshold curve for the CCD imaging system. Finally, we discuss the experimental method to investigate the electromagnetic effects. This work can provide supporting data and referable experience for others’ evaluation and protection of equipment subjected to intensive electromagnetic environment.

We develop an integrated analysis process on time domain comprising models of disturbances, structures, optical systems, and sampled imaging, and realize the end-to-end system integrated simulation of the space camera from micro-vibration disturbance input to imaging output. Accordingly, a data-transforming toolkit is developed in MATLAB for transferring the structural analysis results to the optical analysis software. The integrated time-domain analysis is carried out in the ground micro-vibration tests of a space camera. The analysis results denote that: under load conditions of two different imaging modes, the maximum statistical values of the dynamic image motions of the camera are 0.42 pixel and 0.27 pixel, respectively; the descending factors of the modulation transfer function of images from the sampled-imaging simulations are 0.951 and 0.974, respectively, which have no obviously visual difference compared with that of original images. These results provide a guidance for the effective prediction of the micro-vibration influence of the space camera.

Virtual reality is an important head-mounted display (HMD) technology trend. A large field of view (FOV) virtual reality HMD device is designed based on the principle of FOV stitching for a wide FOV angle and a good sense of immersion. It is realized by stitching four sets of visual optical systems with a same horizontal FOV angle of 80°. This device achieves a binocular transverse FOV angle of 162°, where the transverse FOV angle of each monocular system is 126° and the overlapping FOV angle of the binocular system is 90°.

The radiation temperature measurement system with a wide dynamic range requires different integration time and neutral density attenuators with different transmittance levels for the radiation sources exhibiting different temperatures. The attenuators and integration time often need to be re-calibrated because of their variation. The re-calibration process is cumbersome, and thus the efficiency of the whole system is reduced. First, based on the calibration theory, a radiation calibration model of the radiation temperature measurement system with a wide dynamic range is developed herein. This model considers the integration time and the transmittance of the neutral density attenuators. A simplified calibration method is also proposed to realize the calibration models of attenuators having different radiation transmittance at different integration time. Two different integration time are calibrated using the proposed method. Then, the gray value response of the system caused by internal dark current and stray radiation is calculated by calibrating at integration time of 0.8 ms and 1 ms using an attenuator having 0.0278% transmittance. Therefore, the calibration models of the attenuators having 0.0740% and 0.8193% transmittances at different integration time can be derived. The precision of temperature measurement is experimentally tested. The experimental results indicate that using the proposed calibration method, the maximum temperature measurement errors of the calibration model for the 0.0740% attenuator at integration time of 0.8 ms and 1 ms are ≤0.36% and ≤0.46%, respectively, and that for the 0.8193% attenuator at integration time of 0.2 ms is ≤4.5%. As for the proposed method, within a certain error range, the calibration efficiency is improved and a certain precision of temperature measurement is guaranteed.

An object distance-aided single-station pose information acquisition method is proposed to solve unstable or non-convergence problems existed in minimum-to-long-distance single-station pose measurements on the premise that the object distance of a feature point can be obtained. First, an equivalent method of continuous discretization of object surface is proposed to calculate image length. Second, a single-station perspective pose measurement model is established using image length as the matching element. Then, the algorithm is tested and verified considering barrel as an example. Finally, the error analysis is conduct for the key factors. The results show that the relative root mean square errors of yaw and pitch angles are 0.97° and 0.90°, respectively. The proposed method can be extended to the single-station pose measurement of the non-axisymmetric objects revolving around aircrafts.

We propose a new method to measure two-dimensional particle size of nanoparticles in imagery based on the principle of dynamic light scattering, which is called the translation-rotation image-dynamic light-scattering (TR-IDLS) method. The nanoparticles in Brownian motion in the sample pool are illuminated by a focused polarized Gaussian beam; then, the horizontal and vertical polarized scattering signals of the nanoparticles are recorded. According to the time-correlation function applied to the fluctuation of scattered-light intensity in the two polarization directions, the distributions of translational- and rotational-diffusion coefficients of nanoparticles are calculated, and the aspect-ratio, equivalent-diameter, and thickness distributions are obtained. The spherical standard nanoparticles and mica-flake particles are measured by this method. The shape and equivalent diameter of the mica-flake particles obtained by scanning electron microscopy are compared with the experimental results obtained by the TR-IDLS method, thereby verifying the feasibility of the TR-IDLS method.

In this study, an infrared integrating sphere radiation source with high uniformity and wide dynamic range is designed to satisfy the high-precision calibration and performance testing requirements of infrared remote sensors. The designed radiatioin source adopts a carbon fiber quartz electric heating tube as the infrared radiation medium. The sphere's working band covers 3-15 μm. Based on the cavity radiation theory of integrating spheres and the blackbody radiation theory, the uniform-light mode of child and parent gold-plated integrating spheres is proposed. The proposed mode effectively improves the uniformity of the infrared integrating sphere. A programmable gold-plated aperture is also designed to realize the linear adjustment of the dynamic range. The relationship between temperature variation and radiance output stability and the temperature control accuracy of the child and parent gold-plated integrating spheres are determined. Characteristics of the infrared integrating sphere are analyzed and tested. Results reveal that the uniformity of the infrared integrating sphere in the area of Φ200 mm of the normal at the light exit is 98.87%, the angular uniformity from -15° to 15° in vertical plane is 99.69%, the dynamic range is nearly linearly adjustable, the background radiation is less than blackbody radiation at the same temperature, and the instability is 0.16%. These results demonstrate that the infrared integrating sphere possesses excellent performance. The infrared integrating sphere calibration source is an effective supplement to the traditional blackbody radiation source and has potential application value in the laboratory spectral radiometric calibration of infrared remote sensors.

A general model for calculating the infrared-radiation intensity of extended aerial targets and small infrared targets is proposed. The model eliminates the influence of background bias, which depends heavily on ambient temperature, and can effectively improve the measurement accuracy of infrared-radiation characteristics of aerial targets in complex and changeable environments. The calculation of small-target infrared radiation intensity is discussed based on the characteristics of dynamic infrared test data of a type of aerial target. The results show that the contribution of the apparently atmospheric radiation intensity to the total radiation intensity is negligible for small medium-wave infrared targets. A simplified model used to calculate total radiation intensity is presented, but it is not applicable for small-target long-wave infrared radiation. The dynamic small-target infrared test data and the extended-target data are processed and compared with theoretical results, and the results verify the correctness of the model.

To satisfy the online detection requirements of the ultralow emission of particulate matter in a coal-fired power plant, this paper proposes the online detection method by the light intensity modulation technique based on the laser-scattering method. In addition, the wavelength modulation technique is combined to propose the synchronous measurement method of gas and particle concentrations. Meanwhile, a mathematical model of the scattered-light signal under wavelength modulation is built. The theoretical analysis and simulation results show that the gas concentration is proportional to the height of the normalized second harmonic amplitude peak and that the particulate matter concentration is proportional to the amplitude of the first harmonic of the scattered-light signal from which the gas absorption has been subtracted. A 1392-nm near-infrared laser is used to simultaneously measure the particulate matter and water vapor concentrations during mosquito coil combustion. The experimental results show that the deviation between the measured water vapor concentration value and that by the temperature-humidity sensor is less than 3%. Moreover, the characteristic value of particle concentration has a high linear relationship with the measured value by the dust meter and the fitting factor R2 is 0.9973. The synchronous measurement of particulate matter and gas concentrations is successfully confirmed based on the wavelength modulation technique.

Herein, foreign object detection in the intralipid solution has realized by the cross-correlation peak of chaotic laser. The foreign objects are put inside the intralipid solution. Chaotic laser with delta characteristics of autocorrelation function is used as the detection source and divided into two parts. The first part is used to illuminate the intralipid solution, wherein the detection signal is received and output by the photoelectric detector through the intralipid solution. The second part passes through the photoelectric detector as the reference signal. The reference signal is cross-correlated with the chaotic laser passing through the intralipid solution. The optical informations of the intralipid solution and its internal foreign objects are reflected by the cross-correlation peak of the chaotic laser. Results show that the correlation detection method of chaotic laser has a strong anti-interference.

We analyze and experimentally verify the impact of facet reflectivity on slope efficiency and output power of 450-nm GaN-based semiconductor lasers. The results reveal that for asymmetric resonator structures, the nonlinear effect of longitude spatial hole burning can be suppressed by optimizing the facet reflectivity, thereby improving the differential quantum efficiency and maximum output power of the device. A high slope efficiency of >1.3 W·A -1 is obtained at the facet reflectivity of 5%, and a high power output of 2.6 W is obtained at the operating current of 3 A.

A 10-mW 1064-nm dual-frequency laser is designed herein by using a single-frequency laser with a single non-planar ring oscillator and an acousto-optic modulator. The range of the beat frequency is 125-175 MHz. The 1064-nm dual-frequency laser is amplified to a power of 10 W by a fiber optic power amplifier. The 2.26 W intensity-modulated green laser, which corresponds to the maximum frequency-doubling efficiency of 24.5%, is obtained to improve the second harmonic generation efficiency by using two 15-mm-long MgO∶PPLN crystals. The obtained green beat frequencies are 150 MHz and 300 MHz when the frequency difference of the fundamental light is 150 MHz, and they had a 2-min beat frequency stability of 2.7 Hz and 5.3 Hz, respectively.

Based on the self-developed large mode area Yb-doped double-cladding 25/400 μm active fiber with uniform doping and low thermo-optic coefficient, an all-fiber high-power narrow-linewidth fiber laser amplification experiment was carried out. The laser system achieves a single-mode fiber laser output with a maximum power of 2.2 kW, a linewidth of 25 GHz, and a center wavelength of 1060.3 nm, with a slope efficiency of 78% and a beam quality factor of M2≈1.2. This is the highest output power currently reported for narrow-linewidth single-mode amplifiers based on domestic Yb-doped double-cladding 25/400 μm active fiber, to the best of our knowledge.

A method for enhancement of a single shot multibox detector (SSD) for aerial infrared target detection is proposed. Herein, the relationship between the sensing field and number of feature layers is analyzed, and a bidirectional feature map fusion mechanism that uses both pooling and deconvolution operations is proposed to enhance the feature expression ability. The semantic enhancement branch of the shallow feature map is introduced and the prediction boxes on the high-resolution feature map are increased, so that the positing accuracy of small-size targets is improved. Comparative experiments on the VOC2007 small object dataset and an aerial infrared target dataset reveal that the mean average precisions increase by 7.1% and 8.7%, respectively, accompanied by a slight decrease in detection speed. The results demonstrate that SSD enhancements can achieve good performance in aerial infrared target detection.

For the erroneous detection of feature points and multi-adjustment of the thresholds caused by global thresholds in the extraction of feature points from scattered point clouds, an adaptive feature extraction method based on Morse theory is proposed. Firstly, the potential feature points are marked by assigning weights that are computed by covariance analysis. Secondly, the mean angle along the principle direction between the point and its neighboring points is defined as the local feature descriptor, in order to compute the discrete gradient of the potential feature points. Finally, the Voronoi diagram of each local feature neighborhood is established,the linear interpolation method is utilized to compute the discrete gradient of the vertices of the Thiessen polygonal, and the gradient extreme points are marked as feature points according to the local discrete gradient vector domain. In order to improve the robustness and the anti-noise performance, the discrete gradient computation is performed using multi-scale where the neighborhood scale is used as the scale parameter, and then the features are extracted with multi-scale analysis. The experimental results demonstrate that the proposed method is simple, robust and does not depend on the sharpness of the features; furthermore, it extracts both sharp and blunt features. The results are satisfactory at different levels of noises from 0.03 dB to 0.05 dB, even though some of the features may be missing under 0.05 dB noise level.

The conventional correlation filtering methods are known to demonstrate poor tracking stability for fast moving and fast deforming targets. Therefore, this paper proposes a correlation filter tracking algorithm for adaptive feature selection. First, the basic features are extracted in the candidate regions using a position filter and a color probability model and fused in different weight combinations to obtain multiple fusion features. Then, the credibility of the fusion features is determined and the features with relatively high credibility are selected as the tracking features of the current frame to estimate the candidate position of the target. Finally, if the maximum credibility is less than the credibility threshold, the detector is activated to redetect the target position; otherwise, the candidate position is just the final position. Meanwhile, the target model is updated to ensure the accuracy of target description. The experimental results on the standard OTB50 and OTB100 datasets show that the proposed tracking method has relatively high tracking accuracy and good robustness under the conditions of motion blurring, illumination variation, and fast motion.

Herein, a method of pulmonary nodule recognition based on a three-dimensional (3D) convolution neural network (CNN) is proposed to overcome the problem of false positives in pulmonary nodule detection by traditional computer aided detection systems. First, a traditional two-dimensional CNN is extended to 3D CNN to fully extract the 3D features of pulmonary nodules and enhance the expressive ability of the features. Second, dense connection network and SENet are combined to enhance feature transfer and reuse, and feature weights are adaptively learned by feature recalibration. In addition, focal loss is introduced as the network classification loss to improve the learning of hard examples. The experimental results on the LUNA16 dataset demonstrate that the proposed network model achieves sensitivities of 0.911 and 0.934 at one and four false positives per scan, respectively, and the competition performance metric is up to 0.891, which is better than that of existing mainstream methods.

In this study, we propose a person re-identification model based on view information embedding. In particular, a pose-sensitive embedding (PSE) network structure is optimized based on the perspective towards characteristics of pedestrian images. First, the fusion part of the PSE feature vector is changed from feature fusion into the concatenation of the feature vectors of three view units, which is considerably reasonable for utilizing different view feature spaces. Second, the view units are separated from the shallow blocks-3 of the skeleton network, which improves the difference of the view feature space. Finally, we design a depthwise separable module based on the improved depth separable convolution to extract features of perspective units, preventing the model parameters from being considerably large and improving the network nonlinearity. The results of the validation experiments conducted using the Market1501, Duke-MTMC-reID and MARS datasets demonstrate that the proposed method can achieve a better recognition accuracy when compared with several advanced algorithms.

The large-scale system matrix generated during the reconstruction procedure of multiple excitation points based on fluorescence molecular tomography (FMT) leads to the high computational complexity and long reconstruction time. In order to shorten the reconstruction time and ensure its accuracy, based on the theory of artificial neural network (ANN), we propose a fast reconstruction method for FMT by reducing the dimension of system matrix in this paper. Specifically, the dimension reduction tool is the autoencoder (AE), which is a famous ANN architecture, and during the training of AE, the input matrix data consists of system matrix and surface fluorescence measurement data, then the representation of the previous matrix in the lower dimensional space is obtained by utilizing encoder part of AE. To test the performance of our method, a series numerical simulation experiments are devised, including non-heterogeneous cylinder and digital mouse experiments. Experimental results demonstrate that our method can effectively shorten the time of FMT reconstruction as well as obtain a good reconstruction accuracy.

We investigate the influence of the structural parameters of the total internal reflection photonic crystal fiber (TIR-PCF) on the Brillouin gain spectrum (BGS) characteristics, including the Brillouin gain, number of Brillouin peaks, and peak relative power. Further, the finite element analysis method is used to obtain the optical and acoustic field distributions for analyzing the acousto-optic coupling effect in the TIF-PCF; thus, the corresponding BGS of the PCF can be obtained. In addition, we discuss the effects of the PCF parameters, such as the air-hole layer number, pitch, and diameter, on the BGS. The law of the Brillouin gain and the acoustic mode number is obtained as a function of the air-hole pitch and diameter. Accordingly, we propose a novel PCF structure, whose air-hole diameters gradually increase from the inside to the outside and which exhibits a similar gradient refractive index distribution. A PCF with two peaks in its BGS and a peak power difference of 8 dB is designed and expected to be used in the fiber optic sensing systems based on the Brillouin beat spectrum to increase the signal-to-noise ratio of the sensing system by a factor of 2.5.

This study investigates the electro-optical phase and intensity modulation characteristics of the optically active and electro-optical crystal and defines its π-voltage. For electro-optic crystals with optical activity, the former concept of half-wave voltage cannot accurately describe the periodical characteristics of electro-optic polarization and intensity modulation. Thus, the π-voltage concept is introduced and defined as the modulation voltage required for the elliptical birefringence phase-delay variation of such a crystal to be equal to π. As for an optically active and electro-optical crystal intensity modulator inserted between two polarizers, the optical activity can provide optical bias for electro-optical intensity modulation. However, the modulated light intensity is an even function of the modulation voltage, and an electro-optical switch can be fully used only when the transmission direction of the analyzer is parallel or perpendicular to the polarization direction of the emerging light from the crystal. As for an electro-optical switcher based on such a crystal, the maximum modulation voltage required to achieve a full switching-state transition can be defined as its switching voltage. The π/4-voltage of a bismuth silicate (Bi12SiO20) crystal with dimensions of 6 mm×4 mm×2.9 mm is experimentally measured and found to be approximately 3 kV for a wavelength of 635 nm. Similarly, the concepts of π-stress and π-strain can be defined for the elasto-optic modulator with optical activity.

A robust least squares integration surface reconstruction algorithm that exploits the relative angle difference is proposed for a topography measurement device. The cost function of the classical least squares integration technique is rewritten based on relative angle difference to avoid accumulation of measurement errors. The space and time complexities are O(N2) and O(N3), respectively. Simulation results demonstrate that the robustness of the proposed algorithm is significantly better than that of Zernike’s wavefront reconstruction method and spline-based least squares integration method. Experiments show that the algorithm can be applied to topography measurement with large aperture angle difference method.

To realize a polarization-independent vertical cavity surface emitting laser (VCSEL), the polarization-independent grating and the half-VCSEL are integrated. Based on the rigorous coupled wave analysis, the effects of grating parameters on the reflection characteristics of a polarization-independent two-dimensional grating are analyzed. The 210 nm high-reflection bands of the polarization-independent two-dimensional grating for a grating period of 691-719 nm and a grating width of 408.73-467.60 nm are obtained via analog computation. Then, the polarization-independent two-dimensional grating and the half-VCSEL with a 1.55-μm central wavelength are integrated to develop a polarization-independent wavelength-tunable VCSEL with the 1.55-μm central wavelength. The calculation based on an optical transmission matrix indicates that the wavelength-tuning range of 93 nm can be realized for the polarization-independent wavelength-tunable VCSEL.

In this study, we propose a global illumination algorithm based on the voxel structure. In the proposed algorithm, the voxel structure is used to store the illumination information of a simplified scene, and the indirect illumination effects can be obtained by collecting the illumination information using the voxel cone tracing method. Only the voxel structure of the dynamic scene is updated in each frame, which supports dynamic light sources and avoids unnecessary calculations for updating the static scene. The voxel structure is a directional hierarchical structure generated by anisotropic filtering, which can be compressed and stored in a mipmapped three-dimensional texture, reducing the storage costs. It is possible to calculate the global illumination effects, such as ambient occlusion and soft shadow, by collecting the outgoing radiance and occlusion values stored in the voxel structure using the voxel cone tracing method. The experimental results demonstrate that this approach can obtain various global illumination effects and exhibit a decent rendering efficiency. Furthermore, the single-frame time is less than 33.3 ms in case of complex scenes.

The GS (Gerchberg-Saxton) algorithm has the disadvantages of slow convergence, low accuracy, and easily falling into local minimum values. The GS algorithm is very sensitive to the initial phase. Usually, the closer the estimated initial phase is to its true value, the better the obtained recovery result is. Therefore, in order to obtain the initial phase more close to its true value, we added a non-redundant aperture mask on the pupil of a telescope, and then we processed the interferogram formed by the mask to obtain the initial phase. This method can effectively improve the convergence speed and running accuracy of the GS algorithm. The efficiency of this algorithm was verified via a computer simulation. The simulation results show that the average recovery accuracy of the proposed algorithm is at least nine times that of the GS algorithm.

Investigating the wind-speed profiles in the Martian atmosphere is significant for elucidating the Martian atmospheric environment. The Doppler wind-detection lidar based on the Mach-Zehnder interferometer is more suitable for Mars ground detection than the normal coherent/incoherent Doppler lidars. However, for detecting the frequency shift of the echo signal within the large field of view received by the Doppler lidar telescope, the Mach-Zehnder interferometer must be adapted by the field-widening technology. This study evaluates the effectiveness of two field-widening technologies—one based on a prism and the other based on the ‘cat’s-eye’ optical system—in a Mach-Zehnder interferometer. The prism-based technology proved more advantageous than the cat’s eye system. Next, a Mach-Zehnder interferometer with an optical path difference of 219 mm was designed and constructed. The designed Mach-Zehnder interferometer was injected with a quasi-parallel beam with an 11-mrad field of view, and its transmission spectrum was measured by scanning a mirror driven by a piezoelectric crystal. The maximum interference contrast of the interferometer was 0.87, sufficient for a Doppler lidar. The height dependence of the interference contrast was then analyzed in the earth’s atmospheric environment. Although the interference contrast of the Mach-Zehnder interferometer with a large optical-path difference decreased slightly with the increase of height in the low-altitude atmosphere (below 5 km), the atmospheric wind speed was still detectable with the interferometer.

To reduce the high false alarm rate of the distributed fiber intrusion monitoring system in outdoor complex environment, this study proposes and demonstrates an intrusion event discrimination method based on integrated time/frequency domain feature extraction. First, a vibration fragment segmentation algorithm based on a self-adaptive amplitude threshold is developed to distinguish the vibrating part. On this basis, the average fragment interval feature is extracted. Next, the vibration fragment with the maximum energy is chosen as the research target, and the length and peak-to-average ratio are extracted in the time domain, whose energy distribution in the frequency domain is calculated according to wavelet packet decomposition and an integrated time/frequency domain feature vector is formed. Finally, one-versus-one support vector machine is used to classify four common intrusion events: footsteps of a passerby, bicycle rolling, knocking on the fence, and cutting of an optical cable. The experimental results show that the proposed method recognizes the abovementioned four common intrusion events with an average accuracy of 98.33%, which is much more accurate than the methods that only extract the time or frequency domain features. Moreover, the proposed method is immune to the optical power variation in light path. Thus, the proposed method is helpful to improve the utility of the system.

Jitter is an inherent phenomenon of on-orbit satellite platforms and is a difficult issue concerning high-resolution earth observation satellites. With respect to ZY3-02 platform's jitter, herein, we propose a phase-correlation registration algorithm based on symmetrical energy distribution. Combining the installation relation among charge-coupled devices at different spectral bands in the multi-spectral camera and considering the linearly smoothing motion characteristics of the satellite platform, we obtain a linear-fitted jitter curve using a quadratic function to eliminate system variation. Finally, satellite attitude data filtered from star tracker and gyro data are used to verify the obtained curve. Experimental results demonstrate that the matching accuracy of the proposed algorithm can reach 0.05 pixel using the simulated data, and thus it can effectively assist in the jitter detection of satellite platforms. Using the proposed method, jitter with the amplitude of 0.41″-1.12″ at the frequency of 0.63-0.65 Hz is detected in the ZY3-02 satellite platform for the first time.

This study realizes a portable sensor for dual sensing of oxygen and temperature using a digital color chip with a simultaneous measurement function of three colors, namely, red-green-blue (RGB). An ormosil sol-gel matrix is doped in the sensing film coating on the color chip with three fluorescent dyes. All dyes are excited using a 405-nm light-emitting diode. The emission wavelengths match the three RGB channels of the color chip separately without a spectral overlap. Under various temperatures and oxygen concentrations, the experimental results show that the developed sensor allows a simultaneous monitoring of both temperature and oxygen with good stability and has an excellent linear response in the oxygen concentration range of 0%-30% and the temperature range of 25-75 ℃.

An improved single shot multibox detector (SSD) algorithm is proposed aiming at the problems of slow detection speed of the target proposal based remote sensing image target detection method represented by faster regions with convolutional neural network (R-CNN) and the low performance in small target detection by the SSD algorithm. The algorithm can combine the advantages of the existing detection methods based on target proposal and one-stage target detection to improve the target detection performance. Furthermore, the algorithm replaces the original visual geometry group net with a densely connected network as the backbone network and constructs a feature pyramid between the densely connected modules instead of the original multi-scale feature map. A sample data online acquisition system is designed to verify the accuracy and performance of the proposed algorithm. A sample set of aircraft and playground target is collected as the experimental sample. The network structure stability is verified by training the improved SSD algorithm. Consequently, good results can be achieved without the support of transfer learning. Moreover, the training process is not easy to diverge. By comparing the Faster R-CNN algorithm using ResNet101 as the backbone network and the R-FCN (region-based fully convolutional networks) algorithm, we find that the mean average precision (MAP) of the improved SSD algorithm is 9.13% and 8.48% higher than that of the faster R-CNN and R-FCN algorithms in the test set, respectively. The proposed SSD algorithm improves the MAP in the small target detection by 14.46% and 13.92% compared to the faster R-CNN and R-FCN algorithms, respectively. Detecting a single image takes 71.8 ms, which is 45.7 ms and 7.5 ms less than that of the faster R-CNN and R-FCN algorithms, respectively.

We investigate the polarization transmission characteristic of typical non-spherical particles to solve the problems encountered by these particles in nature. The polarization transmission characteristics of the ellipsoid, cylindrical, and Chebyshev particles while applying the T-matrix algorithm and the difference in the polarization transmission characteristics of the spherical particles are studied. The results denote that for the ellipsoid particles with medium ratio of transverse and longitudinal axis, the degree of polarization (DOP) variation with different shapes is very small and can approximate with that for the Mie scattering method when the scattering angle is less than 60°. In case of the scattering angle being larger than 60°, the DOP is rather changeable with different transverse and longitudinal axis ratios. Further, the difference between the spherical and ellipsoid particles significantly changes with the increasing ratio. The DOP variation in case of the cylinder particles with medium ratio of diameter and height is more stable than that observed in case of the ellipsoid particles but is still not suitable for performing Mie scattering approximation in the side-scattering and back-scattering area. The polarization curves for the extreme-ratio ellipsoid and cylinder particles exhibit a shape similar to that of a bell and has the largest DOP at a scattering angle of 90°. The deformation parameters and the levels of Chebyshev particles have little effect on the polarization characteristic in the forward scattering area but considerable influence in the back-scattering area. The sensitivity also decreases with the increasing level. These results provide theoretical guidance for studying the polarization transmission characteristic of the non-spherical particles and the approximation problem of the spherical particles.

Stray light resulting from grating multistage diffraction influences the imaging quality of the spatial heterodyne Raman spectrometer (SHRS). In this study, the stray light was analyzed and controlled to enhance the precision of the SHRS. The optical transmission theory was employed to study the stray light produced by grating multistage diffraction in the SHRS using the ASAP software; a stray light controlling structure was designed in -2.5°-2.5° field-of-view via baffling, beam stopping, and light trapping. The results indicate that the stray light ratio of multistage diffraction is reduced from 4.996×10-3 to 1.57×10-8. Thus, the developed structure can efficiently control the stray light from grating multistage diffraction and improve the imaging quality of the SHRS.

The geostationary interferometric infrared sounder (GIIRS) on-board FY-4A satellite is an infrared Fourier transform spectrometer. To improve the quantitative application level of observed data, it is necessary to conduct the on-orbit spectral calibration of the GIIRS accurately. Since the spectral positions are determined by the reference laser frequency and the number of the sampling points collected from interferograms for the Fourier transform spectrometer, the key of the spectral calibration is to ensure the stability of the reference laser frequency. In this study, a line-by-line radiative transfer model (LBLRTM) is used to calculate the reference atmospheric absorption spectra. The effective sampling frequency of the laser is determined empirically by comparing the root-mean-square error between the observed and the reference spectra of the detector, achieving the on-orbit high-precision spectral calibration of the detector. This method is used for the on-orbit spectral calibration of the GIIRS on-board FY-4A satellite, clearly demonstrating its application value.

The interaction between the radially polarized vectorial light and GO/AuNRs composite structures is studied to boost the SERS performance. Based on the simulation of the finite-difference time-domain solutions software, the SERS enhancement factor of the GO/single-AuNR composite substrate illuminated by the radially polarized light is up to 108, which is six orders of magnitude greater than that achieved by linearly polarized light under the same conditions. The physical mechanisms for this performance improvement are due to both the electromagnetic enhancement of gold nanorods excited by the radially polarized light and intrinsically chemical enhancement afforded by the GO film. Furthermore, the effects of different GO thicknesses and the number and arrangement of gold nanorods on the properties of SERS with radially polarization illumination are discussed in detail. The performance control of the SERS based on radial-vector light field bestirring the multifunctional substrate has great potential in biochemistry, food safety, and sensor detection.

Microfocus X-ray source is the core component of micro-computed tomography (micro-CT). This study investigates the relationship between the changes in the focal spot size and intensity of the transmission microfocus X-ray source caused by the lateral diffusion of the electron beam in the target. Results show that, if the density distribution of the electron beam follows a Gaussian distribution, the distribution of the X-ray intensity should also be Gaussian. The standard deviation of the X-ray intensity distribution exactly represents the size of the X-ray focal spot. Furthermore, results show that when the energy deposition of the electron beam in the target reaches 60%, the intensity of the X-ray produced by the target reaches the maximum value, correspondingly. With an increase in the target thickness, the focal spot size of the X-ray source gradually increases; conversely, an increase in the acceleration voltage of the electron beam can appropriately reduce the focal spot size of the X-ray source. This study provides theoretical guidance for target material selection and design of transmission microfocus X-ray source.