To study the differences in scattering characteristics caused by differences in aerosol hygroscopicity, hygroscopic growth models for two types of aerosol are built according to thermodynamic principles and Brunauer-Emmett-Teller theory, respectively. Based on these models, the scattering properties of aerosols at relative humidities from 40% to 90% are calculated using the discrete dipole approximation method. It is shown that when hydrophilic particles, e.g., four typical inorganic salts, reach the deliquescent relative humidity, their size increasese xponentially, whereas that of hydrophobic particles (e.g., soot) increases slowly and nonlinearly with an increase in the relative humidity. Results are consistent with experimental measurements. Using established hygroscopic growth models and the measured refractive index-humidity growth law, the relationship between relative humidity, and aerosol particle growth factor is calculated and analyzed for the particle size range of 0.1--1.0 μm. It is found that the scattering growth factors grow exponentially after deliquescence, their values increases by tens of times at 90% relative humidity, and scattering factor growth curve of sodium chloride inorganic salt agrees well with experimental results. However, the maximum scattering growth factor of hydrophobic particles is only 1.07 at the same humidity, which is much smaller than that of hydrophilic particles. These results provide theoretical support for the assessment of the climatic effects of aerosols and for accurate measurement of aerosol concentrations and atmospheric visibility.

In this study, an optical telemetry method based on airborne and vehicular passive differential optical absorption spectroscopy (DOAS) technology to measure the emission flux of atmospheric pollution gas is studied. This method use an airborne imaging spectrometer and vehicular DOAS spectrometer to synchronously observe navigation within an industrial area. The distribution and diffusion trend of NO2 vertical column concentration in this region are detected through spectral inversion. By utilizing the real-time wind field data and attenuation model of NOx in the atmosphere, the NOx attenuation and the proportion of each component of nitrogen oxides are derived, and thus the emission flux of pollution source NOx is obtained. The NOx attenuation in the atmosphere is corrected by using the proposed method. The NOx emission fluxes of the power and steel plant in the industrial park are calculated to be 3.3331×10 24 molecule·s -1 and 2.6138×10 24 molecule·s -1, respectively. The results show a consistency in airborne and vehicular observations. Compared with the emission flux computed by the uncorrected NOx attenuation, the emission flux accuracy of proposed method has significantly improved by 5%--20%. Compared with the vehicle or airborne independent observation method, proposed method combine the advantages of large scanning range for airborne detection and high spectral resolution for vehicle detection, which is conducive in tracking pollution diffusion and can improve the detection accuracy.

In this study, we established a turbidity inversion model for the Rayleigh-corrected reflectance data from the geostationary ocean color imager (GOCI) based on the buoy observation data obtained from the Yangtze estuary and the East China sea. In addition, remote sensing was used to retrieve the turbidity of the Yangtze estuary and the adjacent sea areas. This research results show that the 680 nm band is the most sensitive to turbidity signals and that the inversion effect can be optimally established by model combining multiple bands. The turbidity distributions in the Yangtze estuary and the adjacent sea areas are observed to be high near the shore and low away from the shore. Further, the turbidity initially increases and subsequently decreases from the Yangtze estuary to the Hangzhou bay, and the changing trend is reversed toward the south of the Hangzhou bay within a day. The diurnal variation of the turbidity of water body, which is affected by the ocean dynamics, can be characterized based on the turbidity zone, and the turbidity of the water outside the turbidity zone does not change significantly within days. The turbidity zone exhibits significant seasonal characteristics, showing a tendency of farther in winter and nearer in summer, which is related to ocean currents.

In this paper, for multi-scale target detection, a multi-scale infrared pedestrian detection method based on deep attention mechanism is proposed. The lightweight Darknet53 is adopted as the backbone network for deep convolutional features extracting, and a four-scale feature pyramid network is constructed to classify and localize objects. The detection performance with respect to small-scale objects such as pedestrians is improved by introducing low-level and high-resolution feature maps. Furthermore, an attention module is designed to replace the traditional upsampling block in the feature pyramid network, which generate local saliency map based on convolution feature, thus suppress the feature responses of unrelated areas and highlight the local feature of the image. Finally, the Caltech pedestrian and U-FOV infrared pedestrian datasets are used to execute two-step transfer learning to ensure the generalization of the proposed model and improve the pedestrian features. The results show that the average precision of the proposed method is 93.45% on the U-FOV dataset, which is 26.74 percentage higher than that obtained using YOLOv3, and the minimum pixel size of the pedestrian that can be detected is 6×13. In addition, the qualitative experiment results obtained using the LTIR dataset validate the good generalization of the proposed model, which makes it suitable for multi-scale infrared pedestrian detection.

A test device for focal-plane energy flux density was designed based on a dual-axis tracking trough concentrator system. The focusing loss caused by the positioning error and tracking error of the receiver was studied through theoretical analysis and experimental testing. Furthermore, the optical loss was quantified by the acquisition factor, and laws of the influence of various errors were revealed. The results show that with the increase in the positioning error, the focal-plane width increases, whereas the focal-plane energy flux density decreases and tends to be uniform. Moreover, the focal-plane center migration and optical loss increase with the increase in the tracking error angle. For the dual-axis tracking trough concentrator system used in this experiment, when the receiver aperture is 50 mm, if the acquisition factor is greater than 90%, the positioning error of receiver is required to be between ±1.1% of 455 mm and tracking error angle is required to be less than 0.111°, and the acquisition factor can reach 95%. The acquisition factor is more sensitive to the variation of the tracking error angle. The experimental results are consistent with the theoretical analysis results, verifying the reliability of the test equipment and methods. The function relation of experiment fitting can effectively guide the engineering application.

Digital holographic system is a promising image-forming system, but speckle noise in the coherent light source of digital holographic system adversely affects the quality of holograms. There are some disadvantages in conventional experimental noise reduction or traditional neural network-based noise reduction methods. In order to realize speckle noise reduction in holograms and balance the efficiency of noise reduction, a fast noise reduction algorithm based on convolutional neural network for single hologram is proposed, and the speckle noise dataset is used to train multilevel neural networks. Theoretical analysis and experimental results show that the convolution neural network applied in digital hologram spectrum domain denoising can effectively improve the quality of the hologram, and multilevel speckle noise can be effectively processed by only one hologram. which can save the effective interference fringes of holograms to the maximum extent while maintaining the denoising performance.

We proposed an improved CycleGAN framework for translating short-wavelength infrared facial images and visible-light facial images. Based on the CycleGAN framework, a loss function calculation path was added and a new loss function was designed. A dataset was established, and the model parameters were adjusted based on experiments to improve the translation effect of the proposed model on the facial images. It effectively overcame the differences in images caused by different spectral characteristics so that the images could be easily recognized. The experimental verification was performed with a self-built dataset. The subjective evaluation, FID(Fréchet inception distance), and recognition accuracy were used to compare the proposed framework with several other frameworks. The results show that the improvement of the proposed framework is obvious and the structural features of the original target are better maintained, which effectively improves the observability and recognition accuracy of image translation results.

In this study, we develop a novel depth estimation method of light field image based on multi-cue fusion to solve the problem of occlusion in depth estimation process. Further, we employ the constrained adaptive defocus algorithm and the constrained angular entropy metric algorithm to obtain the defocusing and consistency clues of the scene, respectively. Subsequently, we determine the initial depth and confidence of the scene. The Canny operator is used to extract the edge information of the central perspective image for enhancing the image edge contour information. The initial depth, confidence, and edge information obtained from these scenes are fused using a Markov random field to achieve high-precision depth estimation. When compared with other advanced algorithms, the proposed method is more effective in solving the occlusion issue. The obtained depth map has higher precision, better smoothing effect and better edge retention effect.

This study proposes an optical multiimage authentication method for attenuating crosstalk noise in multiplexed images and addressing information security problems at different authentication levels based on hyper-chaotic amplitude masks and phase information multiplexing in the Gyrator transform domain. According to the proposed method, hyper-chaotic random amplitude masks are constructed first using a fractional hyper-chaotic Rabinovich system. Then, original images are encoded using a modified Gerchberg-Saxton algorithm for low- and high-level authentication. The hyper-chaotic masks are used as the amplitude constraint to iteratively obtain target images. The obtained N target images are then encoded into a composite image. Finally, the composite image is converted into two phase-only masks for transmission using the modified Gerchberg-Saxton algorithm. Results demonstrate that users can have their own authentication keys in the process of authentication at different security levels. In low-level authentication, the correctness of the authentication image can be determined by retrieving the nonlinear correlation peak between the authentication and original images. In high-level authentication, the authentication image can be obtained to be highly similar to the original image with the peak signal-to-noise ratio and correlation coefficient reaching 25.2817 dB and 0.9844, respectively. The proposed method is robust against occlusion and noise attacks and provides a new idea for optical authentication of multiple images at different levels.

In order to improve the detection probability of a single-frame infrared image and stably detect weak and small targets in image sequences, a weak and small infrared target detection algorithm under complex sky background is proposed based on the improved bilateral filtering and polynomial fitting. The background correlation factor is introduced into the weight coefficient of the traditional bilateral filtering algorithm to effectively reduce the influence of the target point during background suppression, improve the signal-to-noise ratio of the target area and the detection rate of the single frame image. In order to further eliminate the false target, the motion features of the fusion targets are combined to perform multi-frame confirmation on the target point. Aiming at the target miss detection caused by target flicker in sequence detection, the polynomial fitting algorithm is introduced to predict the target position of the next frame, which effectively avoids the problem of truncation of the target trajectory. The experimental results show that the algorithm can stably detect weak and small target trajectories under complex sky backgrounds when the signal-to-noise ratio is less than 2.

Existing light pen systems are required to arrange a plurality of measuring spots to overcome the geometric constraints of a measuring terminal. This will result in a complicated measuring terminal structure. Moreover, it is disadvantageous to use the light pen system in limited measurement space. In this work, the light field imaging method is introduced into a light pen type measurement system. The epipolar plane image (EPI) of the light spot on the light pen is obtained using the light field imaging principle, and the spinning parallelogram operator (SPO) is adopted to calculate the slope of diagonal line on EPI. Subsequently, the depth information of the light-emitting point is estimated and the wrong solution of the PNP (perspective-n-points) algorithm is eliminated. Ultimately, accurate three-dimensional coordinate for the measured point are obtained. The effectiveness of the proposed method is verified by experiments. The proposed method only requires a monocular light field camera combined with three-point light pen to complete the measurement of three-dimensional coordinate value. Thus, the operation of the light pen measurement system can be greatly simplified.

An adaptive fringe-pattern projection method based on image fusion and interpolation prediction is proposed. Firstly, based on multi-mask image fusion, the saturation threshold required for the optimal projection gray value is obtained, and the optimal projection gray value is obtained by combining interpolation prediction and search algorithm. Then, by reducing the overall projection intensity, the coordinate matching is carried out in the unsaturated condition, and the adaptive fringe is finally generated. Finally, the generated adaptive fringe is projected to the object to be measured, and the phase solution and three-dimensional shape reconstruction are performed by heterodyne multi-frequency phase shift method. The experimental results show that the phase information in local over-exposure region can be obtained completely by the proposed method, the average error and standard deviation in absolute direction and forward direction are smaller than those obtained by traditional method, and the average error in absolute direction is reduced by 84.1% and the standard deviation in forward direction is reduced by 69.4%. The proposed method effectively solves the difficult problem of 3D shape measurement of high-reflective surfaces.

A new 1.7 μm fiber laser source based on stimulated Raman scattering of hydrogen in hollow photonic crystal fiber is reported. Simple steady-state coupled wave equations containing only pump light and first-order Stokes light are established and simulated. A homemade 1550 nm nanosecond pulse fiber amplifier is used to pump a commercial hollow photonic crystal fiber about 3 m long and filled with high-pressure hydrogen. The rotational stimulated Raman scattering of hydrogen molecules is used to realize the efficient conversion of Stokes waves at 1705 nm. When the air pressure is 1.2 MPa, the maximum average output power is about 0.5 W (the monopulse energy is about 2.5 μJ), and the maximum optic-to-optic conversion efficiency is about 32% (relative to the total pump power). Research results provide an effective new way to realize the output of high-power 1.7 μm-band near-infrared laser.

According to the Fourier heat conduction theory and thermal stress field theory, the COMSOL simulation software and Matlab software are used to construct the combined damage model of three-junction GaAs solar cell under combined laser irradiation, and the temperature and stress field distribution of solar cell under single millisecond laser irradiation and combined pulse laser irradiation are calculated. The results show that, compared with single-millisecond laser, the combined laser irradiation can produce a wider melting damage area and obvious stress damage, and the damage area and depth will increase with the increase of energy density and action time delay of nanosecond pulse laser. When the energy density increases to 0.5 J/cm 2, the radius and the depth of the melting damage spot are risen to 2 mm and 1.5 μm, respectively. When the time delay increases to 0.5 ms, the radius and the depth are increased to 1.4 mm and 1 μm, respectively.

A near-infrared laser carbon dioxide (CO2) sensor system is developed using tunable diode laser absorption spectroscopy (TDLAS). The system consists of a distributed feedback laser with a central wavelength of 1572 nm, a dense spot-type gas chamber, and an indium gallium arsenic detector. LabVIEW program is used to extract the amplitude of the second harmonic signal and the CO2 concentration is inverted. In order to characterize the sensor performance, gas detection experiments are carried out by using the system. The results show that,when the modulation depth is 0.32 cm -1, the amplitude of the extracted second harmonic signal is the largest. In the volume fraction range of 0-3%, the amplitude of the second harmonic signal has a high linearity with the CO2 concentration (the goodness of fit is 0.999). When the volume fraction of CO2 is 0, the retrieved concentration fluctuation range is -2×10 -4-1.17×10 -4 after continuous test for 1 h. When the integral time is 297 s, the lower limit of sensitivity detection of the system is 2.7×10 -6. Considering the gas diffusion time during dynamic gas distribution, the response time of the system is 40-42 s. The CO2 concentration in the atmosphere is measured continuously for 15 h, and the average CO2 volume fraction measured is about (560±46)×10 -6. Compared with the reported sensors, the gas sensing system presents similar quality factors and can be popularized and applied in industrial and agricultural production, environmental protection and other fields.

In order to study the second-order Raman properties of ZnWO4 crystal, a second-order Raman laser based on ZnWO4 crystal is built, and the second-order Stokes laser output of 670 mW at 1318.3 nm with a repetition frequency of 9 kHz is realized. The corresponding pulse width is 3.294 ns, the optical-to-optical conversion efficiency is 4.7%, and the peak power is 22.6 kW. The experiment results show that the ZnWO4 crystal has good performance and can realize second-order Raman laser output.

The development of extreme ultraviolet(EUV) lithographic objective design is toward the direction of an anamorphic magnification objective system, and large field of view and high numerical aperture (NA) for projection objectives are both needed, which cause an extreme increase of incident angle and incident angle range of an objective lens system, so new methods for multilayer film design of anamorphic magnification EUV lithography objective systems are needed to explore. A progressive optimization multilayer film design method is presented to increase the reflectivity but not to change the imaging performance. This method is successfully applied to design the multilayer films of an anamorphic magnification EUV lithography objective system with NA=0.6. The results show that the average reflectivity of each mirror is higher than 65% and the reflectivity peak-to-valley value of each mirror is less than 3.35%, meanwhile, a good uniformity of reflectivity is maintained.

In this study, we propose an inverse reflectance model based on co-located images to precisely model the nonlinear reflection behavior of the non-diffuse reflective surfaces. The proposed model can accurately map the pixel value to the product of the normal vector and the light direction. We need to capture only one co-located image and one RGB image under multispectral conditions to ensure that photometric stereo vision can achieve a high-precision performance, so the time required to capture images is considerably reduced. To perform surface inspection in case of mass production, the proposed method can realize online detection of the moving surfaces at a microsecond shooting rate because the co-located image can be acquired in advance and used for the subsequent workpiece. However, the iterative steps applied in the traditional methods are omitted, and the robustness with respect to outliers, such as shadow points and highlights, is improved, because a neural network is used in the proposed method to train the near-field photometric stereo model. Furthermore, the results of simulation and experiment show that the algorithm can recover the normal vector of the non-diffuse surface well under the condition of very few images.

The problem of the calculation of the tristimulus value when the shape of the instrument function is slightly deviated from the symmetrical triangle is studied, and the corresponding optimal weighting table is explored. The results show that the optimal weight can be obtained by solving the linear equations of three coefficient matrices, and the coefficient matrices are symmetric positive definite tridiagonal matrices. The optimal weighting table method obtained in the previous study is extended to the case where the instrument function shape is asymmetric triangle. The simulation results show that the accuracy of the optimal weighting table method is better than that of the three-point and five-point calibration methods for the 10-nm and 20-nm measurement intervals. For the measurement interval of 5 nm, the best calculation method is to conduct three-point correction of the measurement data first, and then use the direct selection method to calculate the tri-stimulus value.

At present, all stations in the Loran C chain are limited by the accumulation errors of atomic clock and the measurement accuracy of radio signal, and there has not been achieved high-precision time synchronization so far. Based on the continuous variable entanglement swapping method and the delay device located at the main and auxiliary stations, the phase difference information of the orthogonal components of the entangled light field can be detected with high precision to obtain synchronization time difference information of the main station and the multiple of secondary stations. Through theoretical derivation and simulation analysis, it is concluded that the synchronization accuracy of multiple stations can reach the picosecond level. By analyzing the security of the inter-station synchronous quantum channel, it is concluded that any external eavesdropping behavior will lead to the raise of bit error rate of the main and auxiliary stations.

To obtain the moon elevation with high accuracy, it is necessary to accurately estimate the pointing error of laser. During the on-orbit operation of the lunar orbital laser altimeter (LOLA), the laser pointing at night deviates from that during the daytime due to the large temperature difference between day and night. This study first establishes a theoretical model of the detector’s received energy when the spot deviates from the center of the receiving field of view. Then, it analyzes the theoretical relation between the spot’s offset and relative received energy based on this model, and puts forward a laser pointing error estimation method based on spot energy. Subsequently, using the energy data generated when LOLA passed through the Aestuum region during the mapping orbit (12 months), this energy data is combined with the proposed estimation method, and laser pointing errors of LOLA at night in moon are estimated to be 140.62 μrad along the orbit direction and -413.17 μrad in the vertical orbit direction, which are consistent with the results derived from LOLA earth scanning experiment and orbital intersection altitude data.

The performance of focusing evaluation function is an important factor that affects the on-board autofocusing of a scanning remote sensing satellite. Based on the imaging properties of the agile remote sensing satellite and computing characteristics of the on-board processing unit, an autofocusing strategy of the agile remote sensing satellite is designed, and an improved gray gradient function is proposed as the focusing evaluation function. Based on the traditional gray gradient function and the relationship between image edge information and defocus state, the proposed method takes gray standard deviation as the edge threshold to extract the effective edge information and calculate the image definition. The study results show that the proposed method has high sensitivity, strong unimodality and good timeliness, and is robust to the scene change and noise in the process of focusing imaging. Compared with other commonly used gray scale focusing evaluation function, the proposed algorithm shows obvious advantages.

The environmental changes occurring throughout the world have become the focus of global scientific research. The spectroscopic environmental monitoring technology can be used to obtain information about pollutants based on the optical absorption, emission, scattering, and atmospheric radiation transmission principles by establishing a characteristic factor fingerprint spectrum database and a quantitative analysis algorithm. Further, this technology can be applied to dynamically monitor the air quality as well as the fixed and mobile pollutant sources. The environmental spectroscopy monitoring technology, which can perform real-time and rapid monitoring at high sensitivities and large scales, has attracted considerable research attention. To date, several principal environmental monitoring technologies and systems have been established. These mainly include investigations and applications of the lidar technology, differential optical absorption spectroscopy, tunable diode laser spectroscopy, and Fourier-transform infrared spectroscopy. Based on these technologies and systems, the acquisition, transmission and sharing of monitoring information provides basic environmental information for the whole society, which promotes the development of the environmental quality assessment systems based on the monitoring data and provides the scientific basis for environmental management in China.

A clustering-optimized fast independent component analysis (FastICA) de-mixing algorithm is proposed. It solves the problem of unstable de-mixing information caused by the sensitivity to the initial value of the de-mixing matrix for FastICA algorithm during the spectral information de-mixing process. Fuzzy C-means clustering algorithm is used to reduce the spectral characteristics of single pigment spectral information, the most representative clustering result is selected as the initial value of the de-mixing matrix, and the clustering-optimized de-mixing matrix is calculated by FastICA Newton iteration formula to avoid the effect of randomly selecting initial values on de-mixing the spectral information of mixed pigments. The experimental results show that, compared with other algorithms, the average error value of the unmixed results of this algorithm is reduced by 0.57, the average fitness coefficient is 99.67%, and the spectral angle matching distance is reduced by 0.53. The proposed method can increase the stability of the FastICA de-mixing results, and improve the de-mixing precision of the mixed pigment spectral information.

In this study, we propose a nonlinear correction method to deal with the nonlinear response produced by the infrared detectors of the Fourier transform spectrometers when the direct current (DC) signal of the interferogram is not available. Initially, the relative correction factor is derived based on the out-of-band spectrum of the spectrum that is to be corrected, and the consistency correction factor is subsequently calculated by the spectra without correction. The experimental results show that the goodness of the linear-fit curve in case of radiometric calibration can be improved from more than 0.99 to more than 0.9999 after the implementation of the proposed nonlinear correction method; further, the absolute bias of the radiance of each channel is observed to become less than 0.15 mW·m -2·cm·sr -1 after radiometric calibration. When compared with the existing correction methods, the proposed method avoids the demand from the direct-current(DC) signal of the interferogram. However, the proposed method is considerably dependent on the blackbody spectra measured at multiple temperatures during radiometric calibration. After obtaining the consistency correction factor, nonlinear correction can be achieved for any spectrum as long as the detector operates in a stable manner.

Contour rendering is one of the most influential factors with respect to the image quality of a laser projection television (TV). Herein, the just-noticeable differences of contour rendering for a laser projection TV are measured and determined using the two-alternative forced-choice method and the staircase 1-up 2-down method. Further, the influential factors with respect to the just-noticeable differences (JNDs) are analyzed, and the similarities and differences of contour rendering with respect to a laser projection TV and a liquid crystal display (LCD) TV are presented. The results denote that the JNDs are considerably affected by the texture features of the image in case of a laser projection TV. The values expressed by σ of one JND differ considerably in case of a laser projection TV and an LCD TV, and the differences between the adjacent JNDs are observed to be similar in case of both the TVs. Finally, a subjective image quality evaluation model, which is helpful for optimizing the image quality, can be established for a laser projection TV based on the experimental results and JNDs with respect to the image attributes, including the white level, black level, and color saturation.