In order to investigate the influence of different scattering models and effective particle radii on the cirrus cloud optical thickness retrieval, we simulate the reflectance of different effective radii based on General Habit Mixture (GHM), Solid Column (SC) and Aggregates of Solid Column (ASC) under different cirrus cloud optical thicknesses. The influence of different scattering models and effective radii on the retrieval results of the cirrus cloud optical thickness is theoretically analyzed. The cirrus cloud optical thickness look-up tables of GHM, SC and ASC are calculated by radiative transfer model RT3 based on vector doubling-adding method. The cirrus cloud optical thickness is retrieved by the look-up table method based on the POLDER data. It is found that the retrieval results are consistent with the theoretical analysis. The theoretical analysis and retrieval results show that, for different models, the cirrus cloud optical thickness retrieval results are obviously different. Compared with the GHM and SC models, the retrieval results of cirrus cloud optical thickness by the ASC model are closer to POLDER products. In addition, for these models, the retrieval results of the cirrus cloud optical thickness increase with the increase of effective radius. The results of the ASC model are no obvious increase with the increase of the effective radius, and the results of the GHM and SC models increase greatly with the increase of effective radius. Therefore, when the instrument lacks the ability to retrieve the effective radius, it is suggested that the ASC model can be used to retrieve the cirrus cloud optical thickness, so as to reduce the influence of the effective particle radius variation on the retrieval results. This study can help to develop cirrus cloud optical thickness algorithms for GF-5 satellite in China.

The influence of cut-off speed in a Zeeman slower on the loading of blue magneto-optical trap based on the distribution function theory is analyzed and the relationship between the cut-off speed and the number of atoms trapped by the blue magneto-optical trap is investigated. A comparison between the fitting results from the distribution function theory and that from the general theory indicates that the distribution function theory model can better analyze the relationship between the above two. The cold atomic sample for strontium optical clock is prepared by the usage of a multi-turns Zeeman slower for slowing down the thermal atomic beam entering the blue magneto-optical trap. In the optical clock system, the design of the Zeeman slower is optimally designed based on the distribution function theory and thus the work efficiency of the Zeeman slower is improved. This research lays the foundation to achieve a more efficient and compact Zeeman slower and the miniaturization of an optical clock.

A method for preparing the ground and low excited states with controllable populations of He atoms driven by a laser pulse is proposed based on an accurate angular-momentum-dependent model potential. The corresponding laser parameters of wavelength, peak intensity and pulse duration are obtained by the numerical solution of the three-dimensional time-dependent Schr dinger equation. The numerical simulation results show that for the 2s, 2p and 3p states, the populations at the end of laser pulse can all reach 100%, while for the states of 3s and 3d, the maximum populations are only 80% and 86%, respectively, because of their very close energy levels and small ionization probabilities.

The nonsequential double ionization (NSDI) of Mg atoms below the recollision threshold laser power density is investigated based on the classical ensemble model. For a few-cycle linearly polarized laser pulses with a power density of 3.0×1013 W·cm-2, the electron pairs in the correlated-electron momentum distribution at a final state mainly locate in the first, second and fourth quadrants. Moreover, the electron pairs in the first quadrant show an obvious correlation behavior. By analyzing the delay time between the recollision and the double ionization, it is found that the different delay time corresponds to different ionization processes and the delay time plays an important role in the process of the electron emission. For the NSDI events with delay time less than a half cycle, two electrons tend to emit with an opposite direction, however, for those with delay time larger than a half cycle, there exist emission electrons with the same direction.

Based on the microstrip structure chemical vapor deposition (CVD) detectors, a coaxial detector with a single-ended connector has been manufactured using coaxial structure output components combined with isolated direct current circuits. The detector is mainly used for X-ray measurement of laser plasma. The detector uses a cylinder CVD diamond with the diameter of 4 mm. One end of the diamond is plated with a piece of grid electrode and the other end is plated with a piece of disc-shaped electrode. The grid electrode can ensure the normal application of the bias, but also allows X-ray to directly irradiate on the surface of the diamond. The temporal response performance experiment of the detector is tested on short pulse laser equipment. The results show that the rising time of the detector is 61 ps. The temporal performance study of CVD diamond detectors lays the foundation for the optimization of detectors.

In order to solve the problem of bag filter leak detection and location in the industry of cement production industry, a method for detecting and locating the broken bag based on fiber-optic distributed vibration sensing is proposed. Weak vibration signal can be detected as soon as dust air flows into the broken bag because of the high sensitivity of the vibration sensing system, and thus whether the bag is broken can be judged. A simulation is carried out, and the weak leak signal caused by dust air can been identified in the low-frequency range. In the laboratory environment, the maximum signal-to-noise ratio of the built phase-sensitive optical time domain reflectometry (φ-OTDR) is calculated to be 10 dB and the real spatial resolution is 23.7 m. In field trial, the support vector machine (SVM) algorithm is applied to distinguishing the abnormal bag from the normal one. The average recognition accuracy rate is tested to be 97.8%. The combination of φ-OTDR and SVM is capable of detecting and locating bag filter leak effectively.

The performance of free space optical communication (FSO) system will deteriorate with the existence of light scintillating caused by atmospheric turbulence. But the performance of FSO can be improved effectively by means of diversity reception technology. To further analyze its impact on the performance of coherent receiving system, we use binary phase-shift keying (BPSK) modulation and heterodyne detection to establish FSO system model with spatial diversity over Gamma-Gamma atmospheric turbulence channel. Bit-error rate (BER) and outage probabilities (OP) of maximum ratio combining (MRC), selection combining (SC) and equal gain combining (EGC) with different numbers of receiving antennas and turbulence regions are analyzed. In addition, the BER and the OP of MRC, SC and EGC are compared with the performance of the traditional single-antenna receiving system with the same receiving aperture. The results show that spatial diversity based on MRC and EGC can improve the performance of coherent communication system over atmospheric turbulence, and the spatial diversity based on SC can improve the performance of the system only when the average signal to noise rate (SNR) is below a certain threshold.

A new fire detection algorithm is proposed for solving the problems of the existing video image fire detection algorithm, such as serious loss of foreground information, high false alarm rate and weak generalization ability. It mainly consist of two parts including foreground extraction and classification decision. In order to extract more accurate foreground region, an improved ViBe algorithm is applied to obtain the selectively updated motion area. Meanwhile the color features in the motion area are classified with a two-stage classifier composed of random forest and support vector machine. In the classification decision module, two novel kinds of early flame features are suggested to describe the ratio of the inter-frame area overlap rate to the intensity of different sections movement in the flame region, and then combined with the Hu moment feature for training the decision classifier. The experimental results show that the algorithm is more adaptable for practical applications with high accuracy, low false alarm rate, strong generalization ability and short response time.

Integral imaging needs to record the spatial information of three-dimensional (3D) objects from different perspectives, therefore the computation is large and cost a long time by computer-generated method. To solve this problem, we present a fast elemental image array generation method by using window interception. The sampling model is established by simulating the structure of the lenslet array, and the elemental images of the virtual 3D object, corresponding to each virtual lens, are calculated based on optics parameters of the display platform. Then, the window interception is used to generate the elemental image array. The elemental image array with any aperture and any arranging structure can be generated by changing the sampling points and the window function. We build an integral imaging display platform based on LED and design the lenslet arrays with the shape of square, hexagonal and circular aperture. Different 3D models are used to compare the processing time and display effect. Experimental results show that without reducing the 3D image quality, the proposed method is faster when the resolution of elemental image is higher than the sampling rate of the lenslet array.

The influencing factors of the viewing comfort of the stereo images is the key issue in the stereo imaging technology field. Combined with the characteristics of the visual attention mechanism, the influence of saturation factor on visual comfort of stereo image is quantitatively studied in this paper. Firstly, the stereo salient degree map is obtained, after that the fuzzy membership principle and the mask processor are put on it to get the final salient stereo image, and the eye tracker is used to verify the correctness of the salient area of the obtained stereo image. Then the subjective experiments are carried out with improved gradual approximation method, and the matching map and difference map of the comfortable saturation of the salient stereo image are obtained through a large number of experimental data. The experimental results show that the maximum average difference value of the comfortable saturation of the binocular view is 0.29. The verifying experiment shows the comfortable saturation ranges with the proposed method have a good universality. This work not only provides a basis for the study of stereo image comfort, but also provides technical standards for the making of stereo content.

In order to improve the security of drivers, an augmented reality head-up display (AR-HUD) system is designed based on image semantic segmentation. Firstly, we propose an improved single shot multibox detector network for semantic segmentation of road scene images. The front end of the network uses VGG-16 to extract the image features, and the back ends of the network are sampled on the feature maps. Thus, the feature map is segmented. Through the training of the network, the pixel level classification results of the scene objects are obtained, namely, the semantic content information of the environment. Then, with analysis of the relationship among real scene, optical display system, and drivers, the virtual information generated by computer is added to the real scene. In this way, the content is registered into the driver's view to improve the safety of driving. Experimental results show that the accuracy of the semantic segmentation algorithm can reach 77.8%, and image processing time of the algorithm for each frame is 45 ms, in other words, about 22 frame·s-1.

In order to improve the imaging quality of ghost imaging and solve the problem of high distortion factor under low sampling ratio, we propose a compressive sensing ghost imaging method based on neighbor similarity(NSGI). The neighbor similarity embodied in the correlation between image pixels contains abundant information regarding the spatial structure of the object. We analyze the principle of compressive sensing ghost imaging and use the neighbor similarity to evaluate undetected targets. According to the principle of greedy algorithm, we adopt the neighbor similarity to optimize the process of image reconstruction, and set up the threshold value of the correlation coefficient to reduce computation load and improve precision. The simulation and experimental results show that compared with the traditional ghost imaging, NSGI can obtain high-quality images based on a low sampling frequency, which will further facilitate the practical application of ghost imaging.

The modulation signal of inverse synthetic aperture lidar(ISAL) has the characteristics of high frequency and wide bandwidth, making traditional “stop and go” model of inverse synthetic aperture radar unavailable. By establishing accurate imaging model for ISAL, translational error, rotational error, and model error existed in ISAL image formation process are analyzed. When applying translational compensation and rotational compensation separately to traditional ISAL, the residual error existed in translational motion compensation influences the precision of subsequent rotational motion compensation. Jointly compensated imaging algorithm based on the Nelder-Mead simplex method and the quasi-Newton method is proposed. The proposed algorithm iteratively searches for motion parameters of the target based on the Nelder-Mead simplex method, the obtained optimal solution is used as the motion compensation term to eliminate three kinds of errors globally. Then the quasi-Newton method is applied for removing the residual error. Simulation results show that compared with the traditional separate compensation algorithm, the proposed algorithm can estimate motion parameters of target accurately and obtain a better quality of ISAL image.

The dispersed fringe co-phasing error detection technology needs a certain method to analyze and process the acquired dispersed fringe image to obtain the corresponding piston error value. Compared with other dispersed fringe image processing methods, the main peak position extraction method can be applied to the detection of piston error in large range or in small range within one wavelength. However, the detection accuracy of the method is very sensitive to the calibration error in the central line position of the dispersed fringe. To modify the effect of this calibration error, this paper firstly proposes a self-adaptive extraction method for fringe central line position based on the analysis of the influence of the piston error on the diffraction pattern of double rectangle hole. Based on this, combined with the basic principle of the main peak position extraction method, a method of correcting the influence of fringe central line calibration error is proposed. Finally, an experimental light path is built to verify the effectiveness of the proposed correction method. The experimental results show that the proposed method can achieve a measurement accuracy of less than 30 nm in the presence of the central line calibration error, which is greatly improved compared to the accuracy before correction. This method effectively improves the anti-jamming capability and engineering practicality of the main peak position extraction method.

Wavelets are suitable for analyzing non-stationary signals because of their good localized capabilities in both space domain and frequency domain. In traditional wavelet transform profilometry, the scaling of the mother wavelet, which moves the center frequency of the daughter wavelet, can be used to meet local signals for acquiring phase information. It is not accurate enough to extract the local phase information just changing the scale of the complex Morlet wavelet in the local position. To solve this problem, we discuss the characteristics of complex Morlet wavelets with different Gaussian window widths in fringe analysis and propose an improved wavelet processing method. The phase reconstruction results of the proposed method was compared with those of the wavelet processing by employing a cost function. The results show that the proposed method combines merits of the complex Morlet wavelets with different Gaussian window widths, which gives more reliable ridge information. The measurement accuracy of the wavelet transform profilometry is improved by employing optimized wavelet ridge extraction technique because it has stronger noise suppression ability in fringe analysis. The validity of the proposed method is verified by computer simulations and experiments.

Laser ultrasonic can realize the localization and depth detection of microdefects on the metal surface owing to broadband surface waves. In this paper, a critical frequency method for measuring the depth of a microdefect is proposed according to the transmission/reflection threshold phenomenon. Based on the principle of photoelastic excitation, a laser ultrasonic testing platform is built and B-scanning images of aluminum alloy samples and the localization of microdefects are obtained by using the platform. Spectral energy distribution of the transmission and reflection waves of a microdefect is analyzed by combining the wavelet transform, and the threshold value θ0 is measured as 1/4 when the transmission/reflection threshold phenomenon occurs. At last, the depth estimation of a microdefect is realized by the critical frequency method. In addition, the influences of sample material and the distance between the excitation points and the detection points are also analyzed. Experimental results show that the critical frequency method based on the laser ultrasonic can realize the depth detection of microdefects on the metal surface, and the transmission/reflection threshold value is independent of the propagation distance and the sample material.

A high-order aberration measurement technique for immersion lithography projection lens based on multi-polarized illuminations is proposed. Aerial images of different polarized illuminations are collected by linear sampling, and the measurement model is built quickly based on principle component analysis; the high-order aberration of the immersion lithography projection lens is measured accurately. Compared with the high-order aberration measurement method based on a test target with eight angles, the proposed technique can reduce the number of samples, improve the efficiency of sampling, and speed up modeling. The lithographic simulator PROLITH is used to validate the proposed technique and analyze the influence of the illumination types on the accuracy of the high-order aberration measurement. The results show that the proposed technique can retrieve 60 terms of Zernike coefficients (Z5-Z64) with measurement accuracy better than 1.03×10-3λ, and its modeling speed is improved by about 30 times.

The in-band relative spectral responsivity is a basic parameter for the detection and evaluation of the in-band response non-uniformity of a polarization remote sensor. A set of devices to measure the relative spectral responsivity of the polarization remote sensor is assembled, which consists of a laser pumping xenon light source monochromator, a depolarizer, a reference detector and a 45° beam splitter. The depolarizer is utilized to eliminate the output polarization characteristics of the monochromator, and the spectro-polarimetric analyzer (SPOLA) is utilized to measure and verify the depolarization accuracy. The synchronization measurement method of beam splitter is utilized to reduce the instability of the light source and to improve the measurement accuracy and efficiency. By taking 490 nm and 870 nm polarization channel of the atmospheric corrector as an example of application, the experiment of in-band relative spectral responsivity in instrumental level measurements is carried out. The experimental results show that the accuracy of the ratio between measured range value of the center wavelength and the average value of the bandwidth for atmospheric corrector is within 0.25%, which meets the in-band responsivity non-uniformity calibration requirement of less than 0.6%.

Aiming at the problems of low recognition accuracy and slow recognition speed due to the fuzzy image of infrared ship targets, a classification algorithm based on deep convolution neural network (CNN) is proposed. By using the marker-controlled watershed segmentation algorithm, the connected regions in infrared ship image are extracted and the corresponding target positions of the original image are marked and normalized to extract the candidate regions. The improved AlexNet (a deep CNN model) is used for ship targets identification. The extracted candidate regions are sent to the improved AlexNet for feature extraction and prediction to obtain the final detection result. The marker-controlled watershed segmentation method can greatly reduce the number of candidate regions and reduce the classification time of deep CNN. The data of nearly one thousand infrared ship images are obtained by the laboratory-made infrared imaging system, and the simulation experiment on the dataset formed by its translation and scaling is performed. The simulation results show that the combination of the marker-controlled watershed segmentation algorithm and the deep CNN can effectively identify the ship targets. The proposed method has good performance and can identify infrared ship targets more quickly and accurately.

In the view of the problem that there is no high precision universal magnification measurement methods and apparatus at present, a high precision magnification measuring method for imaging system based on double fiber point-diffraction interferometer is proposed. The separation distance between two fibers in object plane and the separation distance between the fibers’ imaging points in image plane are measured with nanometer accuracy by analyzing the quantitative relationship among the separation distance between double point source, spatial location of CCD camera and the Zernike polynomial coefficients of the phase distribution of the point-diffraction interference field. And then, high precision magnification measurement is achieved. Simulation and experimental verification have been carried out to demonstrate the feasibility and stability of the proposed measuring method. The results show that the expanded uncertainty of the magnification measurement is 2.64×10-6. The proposed method has the merits of high accuracy, high efficiency and high reliability. It can be used to measure the magnification of high accuracy imaging system with ultra-high precision, such as microscope objectives and lithographic projection lens.

A Mach-Zehnder (M-Z) electro-optical modulator with low driving voltage and high modulation parameters is designed on the X-cut Lithium Niobate (LN) substrate by using a directional coupler instead of traditional Y-shaped branch structure. The working principle and major factors of directional coupler are analyzed and calculated precisely by the finite element method. The major structural parameters of the electro-optic modulator are optimized. The results show that when the light wave propagates along the directional coupler, the energy is alternated back and forth in the two waveguides with a distribution of sine and cosine functions. At the same position, the coupling length increases rapidly with the increase of the gap. When the electro-optical modulation arm length is 3 cm and the optical wavelength is 1550 nm, the half-wave voltage Vπ obtained by the single-arm modulation is 2.22 V and the modulation parameter S21 is -51.13 dB. With the section analysis of modulation arms, the electric field Ex component, the electric displacement vector Dx component and the optical mode distribution are given, and the integral overlap factor is 0.486. It also shows that the ridge waveguide structure can effectively improve the electro-optic efficiency. When the ridge height is 1 μm, the overlap integral factor is 0.714, and the electro-optical overlap integral factor of it is 46.91% higher than the slab waveguide.

We design a phase function to precisely tune the central lobe size of the Bessel-like beams with the axial distance z. We further deduce the analytical formula of the central lobe size evolution with the propagation distance. Experimentally, we perform the precise control of the central lobe size in the direct space with our phase function loaded on the spatial light modulator. The measured evolutions of the Bessel-like beams with different lobe parameters agree well with the analytical and simulated results, which verifies the reliability of the phase function. Furthermore, the transverse intensity profiles at the longitudinal location of the peak intensity can be well fitted by the Bessel function.

Heat transfer from outer thermal shield to an optical cavity is simplified to a multilevel resistor-capacitor (RC) integrating circuit, which is used to calculate the temperature response of the optical cavity to environmental temperature fluctuations. The temperature response of the optical cavity to the distance between the thermal shield and the optical cavity, the number and the thicknesses of the layers of the thermal shield is discussed based on this method when the mass of thermal shield is fixed. The analysis shows that the temperature response time of the optical cavity can be increased by 2 times as the distance between the thermal shield and the optical cavity is reduced from 40 mm to 5 mm. The temperature sensitivity of the optical cavity to the environmental temperature fluctuations can be reduced by at least one order of magnitude when the number of the layers of the thermal shield is increased from 1 to 3 and the thickness of the inner layer of the thermal shield is maximized. The frequency stability of the frequency-stabilized lasers based on the optical cavity can be improved by the optimized design of the thermal shield of an optical cavity.

The non-iterative method with high precision is proposed for the pose estimation of uncalibrated in-camera parameters in incremental motion structure recovery. Based on the pinhole camera model, the equations which describes uncalibrated cameras pose estimation problems are given. The rotation matrix is parameterized by unit-norm quaternion, and then decomposed according to the four-fold symmetry. To eliminate variables properly and build a bivariate objective function, the least square principle and first order optimal condition is adopted respectively. An efficient solver by using the Gr bner basis technique is developed, so as to obtain optimal solution of rotation matrix and focal length. The complexity of proposed method is O(n). Experimental results prove the efficiency, high precision and numerical stability of the proposed solution.

In order to solve the problem of accurate tracking and scale estimation in videos where targets change their scales, we propose a scale adapted tracking algorithm based on kernelized correlation. Firstly, we establish kernel ridge regression model and construct a two-dimensional kernelized correlation location filter. The center location of target is determined precisely by using fused multi-channel features. Then, the multi-scale samples of target area are obtained and their sizes are reset to the same with the model. By extracting their features and reconstructing to one-dimensional vector, we construct the one-dimensional kernelized scale filter to achieve optimal scale estimation. The experimental results on OTB2013 platform, especially on the scale changing benchmark dataset indicate that the proposed algorithm performs better in precision and success rate in comparison with eight mainstream tracking algorithms. Meanwhile, this algorithm can not only achieve an adapted tracking to the scale changing of target, but also locate its position fast and effectively.

One-dimensional (1D) calibration is one of the most important methods in camera calibration because of its advantages of anti-occlusion and low cost. Firstly, we propose the geometric interpretation of 1D calibration rotating around a fixed point based on the 1D homography matrix. Secondly, it is proved that the cross ratio invariance in projective geometry is equivalent to the 1D homography matrix in the pinhole imaging model. Compared to the cross ratio invarianceof four collinear points, the constraint of 1D homography matrix is more general. Finally, aiming at the 1D calibration under planar motion, the intersection point of two extension lines is calculated by 1D homography matrix. A new convenient method is given to convert planar motion into rotating around a fixed point. The correctness of the proposed algorithm is verified by simulation experiments and real experiments. The experimental results show that the calibration accuracy of the proposed algorithm is improved greatly compared with that of the existing methods.

A novel metamaterial absorber based on a graphene-metal hybrid structure is proposed. The amplitude tunability at single frequency and broadband frequency within the microwave domain is realized via the change of graphene Fermi level by altering the applied voltage. The mechanisms of absorption and amplitude tunability of electromagnetic field are illustrated. The numerical simulation of this metamaterial structure under single frequency band is made. The results show that, when the structural parameters are fixed, the absorption intensity of absorber decreases with the increase of graphene Femi level and the maximum modulation depth reaches 58.6%. When the graphene Fermi level is 0 eV, the center frequency changes with the structural parameters. An array composed of unit structures with different sizes is used to realize the characteristic of the broadband wave absorption based on the multi-absorption-peak superimposed expansion bandwidth principle. The property of amplitude tunability possessed by this broadband absorber is confirmed by numerical simulations.

In this paper, we establish a quantitative model based on BP neural network to detect hemoglobin in serum, and put it into practice of hemoglobin measurement, in the result of which various cancer diseases can be completely distinguished. A double-hidden-layer BP neural network model to quantify hemoglobin is established. The correlation coefficient of the prediction set in the model is 0.9838, and its relative deviation is 2.532%. The hemoglobin concentrations in serum samples of breast cancer patients and leukemia patients are obtained by the BP neural network model. A remarkable difference appears in the concentrations between these two sorts of patients (P<0.001), indicating that the model has potential application in distinguishing tumor diseases.

A high precision correction method of illumination field uniformity for photolithography illumination system is proposed. The correction ability and accuracy are improved by the optimization of the fingers' fore-end shape and arrangement of the finger array uniformity corrector. The simulation results show that the correction accuracy of the finger array uniformity corrector is better than 0.2%, when fingers are staggered arrangement. In addition, the correction accuracy of the finger array uniformity corrector is better than 0.16%, when fingers are chamfered, staggered and double layouts arrangement, which is about twice as high as general finger array uniformity corrector.

A kind of experimental method for quickly forming the Gaussian-shaped removal function is proposed. The corresponding nozzle height range of Gaussian-shaped removal function is obtained by the investigation of the relationship between the material removal appearance and the intersection point of the fluid jet center and the workpiece under the tilt fixed-point incidence mode. As for the definite process parameters, the corresponding nozzle height of Gaussian-shaped removal function is obtained by the addition of the rotary motion of the principal axis. The symmetry of the Gaussian-shaped polishing spot is analyzed and the dynamic removal experiment is conducted. The results show that the dynamic removal process based on Gaussian spots possesses a good stability, which can meet the demand of the ultra-precision optical manufacture and also can confirm the efficiency and reliability of this method.

The head-mounted display system is widely applied in the fields of modern education, medical treatment, and entertainment. The off-axis reflective head-mounted display optical system is a design form of head-mounted display system, and it can not only meet the requirements of miniaturization and compactness, but also realize broadband spectrum imaging without color aberration correction. However, the existing designs of off-axis reflective head-mounted display optical systems cannot compatibly meet the demands of large diameter of exit pupil and small F number together. To solve this problem, we design an off-axis head-mounted display optical system based on two reflective free-form surfaces to realize the requirements of large diameter of exit pupil and small F number. The double-curvature free-form surface and XY polynomial freeform surface are used in our designed head-mounted display optical system, its pupil diameter is 10 mm, F number is 3.0, field of view is 28°, and eye relief is larger than 15 mm. The final imaging performance of the system meets the requirements, and it is better than the current design results.

We design a kind of continuous phase plate (CPP) used in high power laser driver to improve the focus beam quality. The optimal design based on controlling segmented spatial spectrum is modified from traditional G-S algorithm. The improved algorithm chooses a restrained initial phase that is related with the low frequency information of goal focal spot’s spatial spectrum, and the medium frequency and high frequency can be appropriately controlled in iterations. The experimental results show that the root mean square error (RMS) of the focal spot can quickly converge to the target threshold under the effect of improved CPP, which compared with that under the effect of the traditional random continuous initial phase. The improvement of initial phase not only can avoids the G-S algorithm easily produces local optimum value, but also the improved method has a good application for the design of asymmetric focal spot.

A kind of three-dimensional (3D) surface-enhanced Raman scattering (SERS) substrate with a high sensitivity is developed for the rapid detection of the trace polycyclic aromatic hydrocarbons (PAHs) in water. The 3D SERS substrates with a high sensitivity are formed by the combination of the GMA-EDMA porous material with the Au nanoparticles with optimal parameters. The signal intensity of this 3D SERS substrate has an enhancement by one order of magnitude compared with that of only parameter-optimized Au colloid solution SERS substrate, and by about two to three orders of magnitude compared with that of natural Au solution substrate. Moreover, there exists a good repeatability. The relative standard deviation (RSD) of detection is 4.78%-9.27% within substrates and 2.05% between substrates. The SERS spectra of these three kinds of representative PAHs, phenanthrene, pyrene and benzo(k) fluoranthene are detected by the usage of this 3D SERS substrate, and their limits of detection are 9.0×10-10、 2.3×10-10 and 5.9×10-10mol·L-1, respectively. The results show that this detection method possesses a simple operation, a good repeatability, and a high sensitivity, which can be used for the trace detection of PAHs in water.

For traditional optical components, there are some issues in chromatic aberration correction of infrared bands, such as complexity of system structure, great loss of light energy, large mass and so on. Taking the chromatic aberration of 4.8 μm and 10.6 μm infrared wavebands as an example, we study the chromatic aberration correction in the infrared band with the grooved grating surface microstructure. According to the generalized Snell law and the theory of finite difference time domain (FDTD), the surface phase distributions of the microstructures are calculated. The double square cylinders grooved grating microstructure is simulated with FDTD Solution software. The widths of the grooved grating microstructure are designed to be L1=400 nm and L2=950 nm, and the height of the grooved grating is K=500 nm. Double square cylinders grooved grating microstructure samples are prepared by a series of other process technologies of centrifugal coating method, electron beam lithography, and ion etching. The influencing factors of photoresist thickness, exposure image quality, and etching groove type are analyzed. The results show that the phase modulation for ranges of 0 to 1.5π and 0 to 2π can be achieved under the condition of two kinds of wavelengths of 4.8 μm and 10.6 μm by changing L1 and L2. L1=408 nm, L2=944 nm, K=495.32 nm, and surface roughness is 16.32 nm, which are within the allowable range of error. The peak transmittances of two infrared bands of 4.8 μm and 10.6 μm are 71% and 64%, respectively. Using the principle of parallel light measuring longitudinal chromatic aberration, we test the longitudinal chromatic aberration of the two infrared waves reduce by 30%, which proves the effect of the grooved grating microstructure device on the chromatic aberration correction.

As for the investigation of the radiative properties for tungsten inert gas (TIG) welding arcs, the net emission coefficients are calculated for argon plasma in the temperature range of 5000-25000 K. A two-dimensional steady TIG arc-tungsten electrode -water-cooled cooper model is established to obtain the arc radiation intensity distribution, the full arc spectrum, and the radiation power of non-vacuum ultraviolet spectrum. A parameter defined as the ratio of radiation intensity to the sum of radiation intensity and ohmic heat is introduced and through the analysis of the distribution of this coefficient in arcs, it is found that the radiation plays a predominant role in the heat loss from arc body area, while the role of heat transport in the energy balance varies with the arc position.

Continuous variable Einstein-Podolsky-Rosen entangled optical fields can be obtained by a non-degenerate optical parametric amplifier (NOPA) operated below the threshold pump power. The optical coating parameters of optical components as input coupler, output coupler and nonlinear crystal for the two orthogonally polarized beams are different manifested as the transmissivity difference of optical coatings of the output coupler for two orthogonally polarized beams. The influence of the coating parameter difference on entanglement degree is discussed in detail based on the experimental research, which provides references for further improving the entanglement degree of entangled state light fields.

Based on the quantum theory of the intracavity polaritons, the electromagnetically induced transparency phenomenon (EIT) in a cavity is investigated experimentally. By increasing the temperature of the rubidium vapor cell, a substantial increase of the number of rubidium atoms is achieved. When the collective strong coupling condition is satisfied, a weak control field is sufficient for the realization of EIT and significant narrow linewidth. With the further increase of the temperature, the collective coupling effect is enhanced and the cavity linewidth is further narrowed.

An airport detection method based on remote sensing images which combines the cascaded regional proposal network with the detection network is proposed. The regional proposal network is improved to get the airport proposal boxes with a high quality, and the loss function of the detection network is improved to increase the accuracy of the airport detection. In addition, the alternating optimization strategy is adopted to share the convolution layers between the two networks, and thus the airport detection time is greatly shortened. The results show that this proposed method can be used to accurately detect different types of airports under complex backgrounds with a high detection rate, a low false-alarm rate and short average processing time.

Imaging simulation is an essential step in camera development. The aerial linear whiskbroom camera has the characteristics of long distance and large imaging angle, which easily leads to the coordinate projection calculation in the imaging simulation process is iteratively non-convergence. To solve this problem, a new iterative coordinate projection calculation method based on visual vector segmentation is proposed. Firstly, the visual vector is segmented in the iteration window, and then the elevation difference between each segmentation and the corresponding ground point is calculated according to the coordinates of the image point and the exterior orientation element of the scan line. The segmentation of minimum elevation difference is found. Then the iterative window continually built at the segmentation is calculated until the minimum elevation difference is less than the given threshold, thus the image coordinates corresponding to three-dimensional coordinates of the ground point are obtained. Finally, the gray value of the orthographic image is given to the simulated pixel point corresponding to the ground point to generate simulation images. The imaging simulation results of the experimental data from three different terrains show that the accuracy of this imaging simulation method is higher than 0.005 pixel. Its robustness is higher and it can be applied to the imaging simulation of large-angle imaging method.

As for the inhomogeneous smog environment caused by the sedimentation phenomenon of smog, the relations of inhomogeneous environment parameters in the experiment and the simulation are established by the usage of the Monte Carlo simulation procedure and the real-time measurement of the optical depth during the sedimentation time, which solves the delay problem in the measurements of the polarization transmission characteristics and the medium parameters in the traditional inhomogeneous environment and the measurement accuracy is also enhanced. The experimental results show that, with the increase of the concentration of the smog, all the degrees of polarization of the four kinds of polarized lights are constantly decreased; the circularly polarized light has a better polarization-maintaining characteristic with the increase of the concentration of the smog；the longer the wavelength is, the better the polarization characteristic of the circularly polarized light is;the shorter the wavelength is, the better the polarization characteristic of the linearly polarized light is.

Accurate acquisition of the bidirectional reflectance characteristic of surface is essential foundation of radiometric calibration and satellite full-chains imaging simulation. Bidirectional reflectance distribution function (BRDF) is retrieved by measured BRF in field environment. The simulation result shows that the relative difference between the BRF directional coefficients of the same surface is 19% when atmospheric visibility is 15 km and 23 km, respectively. Therefore the true directional reflectance feature can’t be characterized by measured BRF in field environment due to the effect of atmospheric scattering. We propose a BRDF retrieval method, by the synchronous measurement of atmospheric scattering radiation in 2π solid angle. In this method, the residual between measured BRF and value calculated by measurement model is used as the cost function in inversion process, the effect of atmospheric scattering is deducted and the true BRDF parameters of surface are retrieved. The bidirectional reflectance characteristic is inherent attribute of surface and does not change with environment radiation.

Forty-eight samples belonging to four common species of Cyanophyta are studied in order to reduce the chlorophyll a concentration measurement errors caused by unstable fluorescence spectra in vivo. We analyze the effects of algae species, growing period, and growing environment on the photosynthetic pigment composition and fluorescence efficiency by measuring the content of chlorophyll a and phycocyanin, as well as the three-dimensional fluorescence spectra of Cyanophyta in vivo. And we obtain the weight spectra by analyzing the spectral instability under different habitat conditions. The weight fluorescence spectrum of Cyanophyta in vivo is established based on the weighted average method. The measurement results that respectively obtained by the weight spectrum and different normalized spectra are compared, and the comparisons indicate that the weight spectrum can significantly reduce the dependence of fluorescence method on algae species, growing period, and growing environment, and thus increase the accuracy of chlorophyll a concentration of Cyanophyta. The relative error of the weight spectrum is 0.1%-30.4%, with a mean relative error of 12.8%. The maximum relative error of anabaena can be reduced by 104.1% in the weight spectrum.

In order to quickly and accurately identify the source types of coal mine water inrush, we propose a method of constructing a multilayer regularization extreme learning machine (M-RELM) model, which combines the functions of nonlinear feature extraction and classification learning. The fluorescence spectra of water samples are obtained by laser induced fluorescence (LIF) technique as the input of model. The features of fluorescence spectra are extracted by the improved auto encoder (AE) to form the feature space of the model hidden layer. In order to reduce the effect of noise and anomaly of spectra on classification results, the algorithm of extreme learning machine(ELM) is optimized regularly. According to whether the unknown samples are used to construct the training set, the model is optimized regularly by the L2 norm regularization (L2-RELM) or the graph manifold regularization (GM-RELM), which realizes the supervised or semi-supervised classification learning. By propagating between the hidden layers of different functions, M-RELM is constructed, and the integration of pre-training and training is completed. The water inrush samples in Huainan area coal mine as the experimental object, the performance compares with the support vector machine (SVM) and ELM with a single hidden layer. On the samples set containing mixed water, the average testing accuracy of the model can reach more than 94% and the training time is about 0.2 s. On all water samples containing the unknown samples, the testing accuracy of GM-RELM is increased by 2% than L2-RELM. The experimental results show that the M-RELM model is more suitable for the identification requirements of coal mine water inrush.

The Cs-doped ZnO nanorod array (CZO-NRA) is fabricated by using the chemical bath deposition technique, which is used as the electron transporting layer (ETL). The mixed solution of ethanolamine and 2-methoxyethanol is used to modify the surface of CZO-NRA, and an inverted polymer solar cell (IPSC) is fabricated. The research results show that, the moderate Cs-doping increases the preferred orientation degree of crystallinity along the c-axis for the nanorod, decreases the deep level defect in ETL induced by the oxygen vacancies and the interstitial Zn atoms, decreases the series resistance of devices, and increases the short-circuit current and the filling factor of devices. The surface modification reduces the surface defects, decreases the contact resistance between ETL and the active layer, and suppresses the interfacial carrier recombination. Compared with that of the undoped devices, the power conversion efficiency of the surface-modified CZO-NRA devices is increased from 1.27% to 2.89%.

The influence of deposition process on the reliability of YbF3 films is studied from four aspects including deposition method, film thickness, substrate temperature and ion beam energy. The research results show that, in contrast to those by resistance heating evaporation, the films deposited by electron beam evaporation have a higher density and lower water-vapor absorption. When the film is too thick or the deposition temperature is too high, the stress of YbF3 film increases, which results in the occurrence of cracks on the film surface or the peeling of coatings. Ion-assisted deposition can enhance the adhesion of YbF3 films and improve the surface quality. As the ion beam energy increases, the stress of coatings first increases and then decreases. According to the above results, the optimal deposition process is obtained and a broadband anti-reflected coating with good optical properties and high reliability is produced.

We propose a fully autonomous micro aerial vehicle with onboard sensors to achieve simultaneous three-dimensional localization and dense reconstruction. Based on the ORB-SLAM system, a visual-inertial simultaneous localization and mapping system is proposed based on the extended Kalman filter, which improves the robustness and accuracy of the system to meet the requirements of micro aerial vehicle autonomous flight. Since sparse feature point maps created by the ORB-SLAM system can't be used for micro aerial vehicle obstacle avoidance and navigation, a stereo camera is used to propose an improved method of building maps from sparse maps to dense octree maps. The experiment evaluation with EuRoC dataset shows that the proposed algorithm improves the precision of open keyframe-based visual-inertial algorithm by one time. The proposed algorithm is applied to the quadrotor autonomous flight platform, and the fully autonomous flight and dense map construction is achieved by relying on on-board sensors and processors. The effectiveness and robustness of the proposed algorithm are verified.