The atmospheric water vapor column density is retrieved based on near infrared passive differential optical absorption spectroscopy technique(IR-DOAS). The high-resolution cross sections are obtained from the Hirtran database and Voigt profiles are used for linear broadening under different temperature and pressure conditions to obtained different inversion absorption cross section. The spectra with elevation angle of 90° is selected as reference spectra to retrieve H2O column to obtain the vertical column density. Compared with solar photometer(CE-318), it is found that the results are in good trend consistency with IR-DOAS, and the linear correlation coefficient is 0.99. The differences between inversion values of IR-DOAS and CE-318 are within the range of inversion error of IR-DOAS. It is found that the slant column density of water vapor in vertical direction varies gradiently with elevation angle, while the slant column density of water vapor in horizontal direction nearly uniform with observation azimuth, and the distribution is uniform.

Under the background of the short-wave infrared imaging spectroscopy application, we study a set of corrections of HgCdTe short-wave infrared focal plane arrays, including defective pixel correction and non-uniformity correction, and propose the correction principle of non-uniformity correction after defective pixel correction. Under standard radiation sources, the normal distribution of normal pixel output values is fitted, and the threshold of normal pixel output is set by the 3σ criterion to determine the number and the location of defective pixels in the detector. Then according to application requirements of short-wave infrared imaging spectroscopy, the spectral two-neighborhood mean replacement is applied to the defective pixels. After the defective pixel correction is completed, the non-uniformity correction of the detector is carried out by the two-point method with small calculation amount and strong real-time performance. The comprehensive correction results show that the defective pixels of the detector are effectively eliminated, the output values of defective pixels are well corrected, the effect of non-uniformity correction is obvious, and the image details are more abundant.

A camera array system composed of commercial cameras is applied in the initial orbit determination of space targets. The overview of the camera array system is introduced. Based on the technical characteristics of the camera array, the feasibility for the application of the camera array in the initial orbit determination is analyzed from three aspects: detection capability of space targets, positioning accuracy and improvement of initial orbit determination algorithm. The main process and the experimental verification are provided. The experimental results preliminarily demonstrate the ability for the camera array system to improve the accuracy of initial orbit determination and the detection magnitude of limit exploratory.

An asymmetrically clipped optical (ACO) multi-carrier (MC) code division multiple access (CDMA) system is proposed based on CDMA and ACO orthogonal frequency division multiplexing. The theoretical expression of signal-to-noise ratio for the system affected by clipping noises is derived with the orthogonality restoring combining (ORC), maximum ratio combining (MRC) and equal gain combining (EGC) algorithms in the sight and sattering propagation channels. The model for Monte Carlo bit-error-ratio (BER) simulation is established. The reasearch results show that the BER performance can be improved with an increase in the length of spreading codes. As the number of users increases, the MRC algorithm deteriorates due to the muli-user disturbance, however the ORC algorithm maintains the orthogonality among different users and demonstrates the best BER performance. The proposed system outperforms the Flip-MC-CDMA and the unipolar multi-carrier CDMA (U-MC-CDMA) systems.

A method for the recognition of fiber-optic perimeter vibration signals is proposed based on local mean decomposition (LMD) and serial feature fusion (SFF), in which the effect of noise is first suppressed to extract the relevant information of vibration signals, then the SFF is conducted to get the feature vectors with the ability of accurate description, and finally the probabilistic neural network (PNN) algorithm is adopted for learning and classification. The proposed method is validated by different single-vibration signals and vibration signals under the stormy weather interference. The results show that, by the proposed method in the above two cases, the average correct-recognition rates reach 96.0% and 96.7%, and the recognition time is 0.87 s and 0.91 s, respectively. The proposed method is superior to the traditional LMD algorithm and the SFF-PNN algorithm in the sensitive information recognition and feature extraction.

Aiming at the problem that the inter-satellite laser communication system is easily affected by the pointing errors, the method of third-order central moment is used to make the pointing error equivalent to the modified Rayleigh distribution under the condition that the pointing errors obey the Beckmann distribution. Under the assumptions that the outage probability and the transmission power are fixed, the optimal model of transmitter power and the optimal model of outage probability are established, respectively. The optimal root mean square width of the Gaussian beam emitted by the antenna under the above two assumptions is calculated. Through the numerical simulation, the numerical relationship of the outage probability and the minimum transmitting power with the optimal root mean square width is given. According to the numerical analysis results, the optimal root mean square width of a Gaussian beam can be selected based on the definite pointing error distributions, so that the optimal link performance for the inter-satellite laser communications is obtained.

A dispersion measurement prototype is designed based on the modulation phase shift method, which improves the accuracy and reduces the cost of dispersion measurement. In addition, the dual phase discrimination circuit is designed based on the amplitude and phase detection chip AD8302 and the accurate measurement of phase difference is realized. The least square method is used to fit the received data and a reliable dispersion coefficient curve is obtained. The modular design is adopted as a whole. The digital signal processing based on chip F2812 is also used to accomplish the control to the board of each module as well as the collection and processing of data. The interface is designed on the LabVIEW development platform. The performance of the prototype is verified by experiments and the results show that the uncertainty of the cumulative dispersion measurement at wavelength of 1550 nm is less than 10 ps/nm for the G.652 fibers with different lengths.

In order to further improve the light energy collected by the photo-sensitive area, a new design method for the aspheric microlens fabricated by the photoresist reflow method is proposed, which combines finite element ray tracing with the Lagrange multiplier method. Given the focal length, the needed aspheric surfaces can be computed automatically by the proposed method. In order to verify the accuracy of this algorithm, the RMS radius of the focal spot of the microlens with circular aperture is simulated and theoretically calculated, respectively, and both results are consistent. The approximation error of spot diagram of the cubic finite element ray tracing is estimated. The enclosed energy collected by the photo-sensitive area of 320 pixel×256 pixel ultraviolet photodetectors is computed and the efficiencies of microlens with square aperture shapes and different fillet radii are compared. The results show that the square aperture microlens with a 6 μm fillet radius has a relatively high energy concentration with an efficiency of 96%.

An airplane detection method is proposed based on feature fusion and soft decision, in which the region-based convolutional neural network is used as the basic framework and the L2 normalization, feature connection, scaling, and dimensionality reduction are in turn used to fuse the multi-layer features. The soft decision, which can improve the traditional non-maximum suppression method, is introduced in order to reduce the detection-omission-rate of grids in the case of significant overlap of targets. The experimental results show that the proposed method can be used to detect airplanes accurately and quickly with a detection rate of 94.25%, a false alarm rate of 5.5%, and the average running time of 0.16 s. Compared with those of the other existing detection methods, each index of the proposed method is significantly improved.

An indoor human detection dataset (IHDD) is established, and a novel multi-view indoor human detection neural network (MVNN) based on joint learning is proposed. The model consists of input data layer, feature extraction layer, deformation layer, visibility reasoning layer and classification layer, and the proposed MVNN algorithm can improve the detection performance when combined with the region proposal model and the multi-view model. Experimental results on the self-built IHDD show that compared with other existing detection algorithms, the proposed MVNN algorithm has a higher detection rate. It can still obtain good detection results even in the case of difficult situations such as various views, changing poses and occlusion for human targets, which indicates certain theoretical research value and practical value.

An accelerated image super-resolution reconstruction algorithm based on deep residual network is proposed to solve some existing problems, such as few convolutional layers, simple model, large amount of calculation, slow convergence speed and fuzzy image texture. This method improves image resolution and accelerates convergence speed at the same time. First, a deep convolutional neural network model is proposed to improve accuracy, and accelerate convergence of network models is achieved by residual learning and Adam optimization method. Second, feature mapping is performed directly on the original low-resolution image, and the sub-pixel convolutional layer is introduced at the end of the network to rearrange the pixels, so a high-resolution image is obtained. Experimental results show that the proposed algorithm has higher peak signal-to-noise ratio and structural similarity index than those of existing algorithms on set 5, set 14 and BSD100 test sets, and can recover more image details. The image edges are complete and the convergence speed is fast.

Aiming at the problem of the severe image degradation under a low-light condition, a low-light image enhancement algorithm based on deep convolutional neural network (DCNN) is proposed. The training sample is synthesized by this algorithm according to the Retinex model. Then, the original low-light image is converted from RGB (Red Green Blue) space to HSI (Hue Saturation Intensity) color space. The luminance component is enhanced by using the DCNN while keeping the chrominance component and the saturation component unchanged. Finally, the image is turned back to the RGB space from HSI color space to get the finally enhanced image. The experimental results show that, compared with the existing excellent image enhancement algorithms, the proposed algorithm can not only effectively enhance the brightness and the contrast, but also can avoid the color distortion and the over-enhancement, and both the subjective vision and objective evaluation index are further improved.

A polarization axis detection algorithm is proposed based on the symmetric features of end-view images, in which the polar coordinate transformation of the images is first used to realize the symmetry detection of images, the pyramid searching method is then utilized to quickly find the optimal symmetric axis, and finally the direction of the polarization axis is determined. Since the proposed algorithm can achieve an accurate positioning by utilizing the global features of images instead of the edge points in the stress region, the requirement for the sharpness of image is greatly reduced. The experimental results show that the proposed algorithm possesses high accuracy and precision (±0.1°), indicating the robustness is greatly enhanced, and simultaneously the speed of this algorithm is improved by nearly 1.5 times.

A method for retinal vessel segmentation is proposed based on a fully convolutional neural network with multi-scale feature fusion, which does not need hand-crafted features or specific post-processing. The architecture of skip connection is utilized, which combines the high-level semantic information with the low-level features. Residual block has been introduced to help learn details and texture features. The multi-scale spatial pyramid pooling module is built by atrous convolutions with different atrous rates to further enlarge the receptive fields and fully combine the context information. The class-balanced loss function is applied to solve the problem of imbalanced distribution of samples. The experimental results show that in the two datasets of digital retinal images for vessel extraction (DRIVE) and structured analysis of the retina (STARE), the accuracies are 95.46% and 96.84%, the sensitivities are 80.53% and 82.99%, the specificities are 97.67% and 97.94%, and the areas under receiver operating characteristic (ROC) curve are 97.71% and 98.17%, respectively. The proposed method is superior to the other existing methods.

A Mojette transform tomographic reconstruction algorithm based on algebraic iteration is proposed. In the algorithm, the optimal projection angles are determined, and the multiplicative algebraic iterative algorithm in traditional tomography is adopted for reconstruction. The proposed algorithm is used to tomographic reconstruction of an axisymmetric flame. Numerical simulations show that, for the projections with high noise, the proposed algorithm has better reconstruction quality comparing with the Corner Based Inversion.

A phase unwrapping method used in the spectral domain phase microscopy (SDPM) is proposed. A wrapped phase with a small noise and an unwrapped phase with a big noise are obtained by using respectively the Fourier transform (FT) method and the synthetic-wavelength phase calculation method. The wrapped number of the wrapped phase is calculated from the difference between the unwrapped and wrapped phases, and thus the unwrapping of the wrapped phase with a small noise is conducted. The presented method eliminates the boundary segmentation error introduced in the existing phase unwrapping methods. An SDPM system based on synthetic-wavelength is established. A piezoelectric translation stage is used to quantitatively verify that this system is capable of accurately unwrapping the phase with a large gradient boundary. The phase imaging of red blood cells and tilted mirror surfaces is also performed. The displacement sensitivity of this system in air is 0.043 nm.

Bioluminescence tomography (BLT) is a non-invasive, highly sensitive optical molecular imaging technique, which can reveal the three-dimensional (3D) distribution of the bioluminescent sources inside the tissue through the light signals detected on the surface. The BLT reconstruction problem is ill-posed due to the domination of scattering during the light propagation through the tissue, which results in a challenge to accurately reveal the 3D source distribution. According to the sparse distribution of the bioluminescent sources, the sparse regularization method based on L1 norm has achieved a significant improvement comparing to the traditional L2 norm regularization. Furthermore, due to the spatial aggregation characteristics of the bioluminescent light sources, adopting this feature would further improve the BLT reconstruction accuracy. Comparing to the traditional sparse reconstruction algorithm which takes all unknowns in the solution domain into account, the feasibility of block sparse priori information used for the BLT reconstruction is explored. First, the solution domain is divided into a series of data blocks through analyzing the correlation coefficient between the columns of the system matrix. Then, the block sparse Bayes learning algorithm is used to reconstruct the distribution of the bioluminescent sources. Through the simulation experiment and the mouse in vivo experiment, and compared with those by the traditional sparse reconstruction algorithm based on L1-LS, the results show that the proposed method can effectively alleviate the ill-posedness of the BLT reconstruction problem, suppress noise, and improve the reconstruction accuracy.

The effect of filter on the column density of polluted gas in the non-dispersive visible imaging system is studied. The relationship between central wavelength and incident angle is analyzed for the filter, and it is found that the central wavelength of the filter shifts toward the short-wave direction with the increase of the incident angle. When the incident angle reaches up to 40°, the center wavelength of the filter has a drift of 17.4 nm. The influences of different filters on the signal to noise ratio (SNR), linear response and sensitivity are also discussed. The research results show that the SNR can be increased by the increase of exposure time or by the image stack. The filter with a full width at half maximum (FWHM) of 10 nm possesses good sensitivity. At the same time, the linear response coefficient above 0.9 indicates that a filter with a FWHM of 40 nm can still satisfy the theoretical condition for resolving the column density of polluted gas in the non-dispersive imaging system. The detection limit for a filter with a FWHM of 10 nm is the best, which is about 4.475×1016 molecule/cm2. The optimal FWHM is between 2 nm and 40 nm for obtaining the column density of polluted gas. Based on the filter with a FWHM of 10 nm and a central wavelength of 450 nm, the two-dimensional spatial distribution map of NO2 column density is obtained by measuring remotely the smoke plume from the chimneys in one steel plant.

The imaginary part of the complex permittivity of the metal nanoparticle is modified by Kawabata-Kubo (K-K) correction. The optical scattering and absorption characteristics of surface plasmons are quantitatively described. The extinction characteristics of light incident on a single ellipsoid metal nanoparticle are calculated and analyzed by Mie theory and the electric dipole theory. A optical polarization structure model of periodic array of ellipsoidal nanoparticles is established. The output characteristics of polarized light from visible to near infrared bands are simulated and calculated by COMSOL software. K-K correction is applied to finite element simulation of the metal ellipsoid arrays by replacing spherical radius with effective radius of ellipsoid. The modified metal nano-ellipsoid arrays have lower transmittance and wider extinction spectrum bandwidth, which is consistent with the experimental trend of broadband strong absorption characteristics of single metal particles.

As a wide spectral range and high resolution infrared spectroradiometer based on linear variable filters has been developed, a new sub-region multi-point calibration method is proposed. Its calibration principle is as follows. First, divide the target temperature region into n sub-regions, measure and record the infrared spectroradiometer data corresponding to n+1 blackbodies with different temperatures in the target region, and calculate the calibration coefficient for each sub-region. During the measurement of an unknown radiation spectrum, compare the blackbody spectra previously measured at calibration temperatures in order to find the proper sub-region. Use the corresponding calibration coefficient for calibration in order to improve measurement accuracy. We use this calibration method to calibrate a self-developed linear variable filter-based spectroradiometer and inverse the blackbody temperature error according to calibration results. The experiment results show that the calibration accuracy better than 1.5 K is achieved, and the proposed method can be used for radiometric calibration of infrared spectroradiometers.

The method for the fast, accurate and reliable extraction of laser stripe centers in the train-running environments is studied. The multi-section fast segmentation of laser stripe is achieved based on the ENet deep learning model. The histograms along the gradient direction of light stripes in each section are counted to determine the normal dominant direction and construct the corresponding direction template. The gray-level centroid method with sub-region multi-template matching is proposed for the extraction of sub-pixel coordinates of light stripe center. The research results show that the proposed method can effectively overcome the influences of various types of interference information on laser stripe center extraction in the outdoor train-running environment. The light stripe extraction time is only 2.1 ms for a single rail profile, the mean error is about 0.082 pixel and the standard deviation is 0.047 pixel. The method has good time-effectiveness and high accuracy for laser stripe center extraction.

A single-baffle metal-dielectric-metal (MDM) waveguide coupled disk cavity cascade structure is proposed based on the transmission characteristics and photon local characteristics of the surface-plasmon-based sub-wavelength structures. The discrete state provided by the side-coupled disk cavity cascade and is used to provide a continuous state produced by the baffle plate placed in metal-dielectric-metal waveguide interferes constructively or destructively, and thus two different modes of Fano resonance are formed. The transmission characteristics of the formed Fano resonance are analyzed according to the coupled mode theory. The structure is simulated by the finite element analysis method and the effects of the structural parameters on the refractive index sensing characteristics are quantitatively analyzed. The effects of temperature and humidity on the measurement results in the actual measurement process are analyzed, based on the physical mechanism underlying the refractive index change. Moreover, the problem of cross-sensitivity in the sensing process is effectively solved by the differential sensing method.

A speckle pattern defletometry (SPD) based on digital image correlation (DIC) is proposed. The two-dimensional DIC is used to measure the three-dimensional shape of a specular surface. The displacements of speckle patterns instead of the deformations of fringes are used to make the deflectometry measurement process simpler and more efficient. Only two speckle patterns need to be taken and the measurement accuracy of surface can reach micrometer level. The principle and method of SPD are described. The relevant formulas are derived, and the design and production of a speckle pattern are involved. An acrylic plastic plate is measured in the experiment. Compared with the experimental results of the phase measuring deflectometry, the measurement accuracy of nearly 1 μm is achieved.

A method based on multi-sensor to measure the pose of a three-dimensional object is proposed. We employ multi-sensor technology to make the advantages of the depth camera and the high-resolution charge-couple device (CCD) camera, improving the robustness and efficiency of measurement. The target area is coarsely located in the point cloud based on the relationship between the object and its fixed board, and the region is converted to gray image space by the prior calibration information. In the gray image, four target straight lines are extracted by the Line Segment Detector (LSD) and feature constraints. The perspective 4 point (P4P) algorithm is utilized to calculate the six-dimensional pose of the object. The experiment verifies the validity of the algorithm and the measurement efficiency of the algorithm is much better than that of the classical template matching method.

In order to solve the underdetermined equation group of multiple spectral channels in a spectral pyrometer, the idea of optimization is introduced, and the problem of solving multispectral true temperature is transformed into a multi-objective minimization optimization (MMO) problem. The true temperature can be solved without assuming the function relationship model between spectral emissivity and other physical quantities. The inversion accuracy of the proposed method is equal to that of the second measurement method, and the inversion speed is greatly improved. With the aid of the actual measurement data of rocket exhaust plume, the MMO method is used to realize the inversion of the true temperature of the rocket exhaust plume.

An improved solution algorithm of the ill-posed problem is proposed to measure the particle size distribution, which is combined with the truncated singular value decomposition(TSVD) method, the modified singular value decomposition method, the Tikhonov regularization method, and the Chahine iteration method. The singular cutoff value is determined by the Backus-Gilbert tradeoff criteria and the minimum principle of singular value. The optimal regularization parameters are determined by the L-curve method, and the simultaneous iterative reconstruction technique (SIRT) is adopted to realize the non-negative constraint of the solution. The simulation and experimental results show that the measurement errors of the single-peak and bimodal distributions are both less than 3% by the proposed algorithm. In addition, the proposed algorithm has obvious advantages superior to the other inversion algorithms in the anti-noise performance, measurement accuracy, timeliness, and measurement range of the particle size.

A miniature underwater shock wave pressure sensor based on optical fiber Fizeau cavity is described. The basic theory and fabrication process of this sensor are given. This transient interference phase of the Fizeau cavity acted by shock pressure wave is demodulated by using passive homodyne demodulation technique. A static calibration test and a dynamic calibration test are conducted using a piston-type pressure calibration machine and a focusing-type electro-magnetic shock wave source. The results indicate that the linearity, repeatability, hysteresis, and intrinsic error of the sensor within the full pressure range of 0-60 MPa are 3.26%, 0.01153%, 0.07%, and 3.407%, respectively. The dynamic response time is less than 0.75 μs.

The integrated waveguide optical delay network for four-element sub-phased array antenna at Ka band (30 GHz) is designed, in which, the radio frequency (RF) signal is converted into the optical domain by the phase modulation and the waveguide micro-rings operate at the anti-resonance state to achieve large bandwidth and continuously tunable time delay. Only one sideband is delayed via the bandpass filtering and the RF signal is recovered by the differential balanced detector. The structural parameters of the cascaded dual waveguide micro-rings are optimized to achieve the continuously tunable time delay in the range from 0 to 24.9 ps, and the bandwidth is larger than 4 GHz. The beam scanning with the maximum scaning angles of ±30° is realized. The gain and noise coefficients of the optical time delay network are analyzed and deduced, and the performance of the whole delay chip system in the practical application is evaluated.

The influences of the concentration of HF molecular and the consumption of the working medium on the laser pulse energy are studied. The laser pulse energy decreases obviously with the increasing of the concentration of HF molecular in the laser, due to the reabsorption of the HF molecule on laser and the strong relaxation effect of the HF molecule on the excited molecules. The laser pulse energy decreases about 1.15% with the increasing of the concentration of HF molecular for 1015 cm-3. In addition, each excited HF molecule can emit 0.8 photons, the dissociation rate of SF6 molecules in the discharge region is about 1%, and the consumption of the working medium is small for a single discharge process and the consumption is about 1/(2×105) of the total gas amount. The experimental results show that the laser pulse energy is affected seriously by the HF molecular concentration, but weakly by the medium consumption. With the add of the molecular sieves in the laser chamber, the HF molecular concentration is maintained at a level of 1.8×1015 cm-3, and the laser pulse energy decreases about 10% under the coaction of two factors.

Based on the spin flip model (SFM), the bandwidth characteristics of the polarized chaotic signals generated subject to dual chaotic optical injection are studied theoretically. The research results show that, in contrast with the case of single chaotic optical injection, the dual chaotic optical injection weakens the injection locking effect, and the vertical-cavity surface-emitting laser (VCSEL) can achieve wideband polarized chaotic signals in a wider range of the parameter space of injection intensity and frequency detuning. Under the given injection intensity, when two frequency detunings associated with the dual chaotic optical injection possess large absolute values, the system is more inclined to produce wide-bandwidth polarized chaos. Under the given frequency detuning, the specific region in the parameter space of two injection intensities required for generating wideband chaos can be determined and expanded if a positive frequency detuning is chosen.

The theoretical research on the Yb…YAG laser pumped by the zero-phonon line with pumping wavelength of 969 nm is constructed, and the rate equations for this 969 nm pumped Yb…YAG laser are established. The optical-to-optical efficiency and the output laser intensity of this Yb…YAG slab laser amplifier pumped at 969 nm or 941 nm are obtained by numerical simulation under the same thermal load state. The simulation results show that the optical-to-optical efficiencies for both are almost equivalent. The pumping intensity for 969 nm pumping is 20% higher than that for 941 nm pumping.

A tracking algorithm is proposed based on correlation filter fusing with keypoint matching. The object is tracked and verified by multiple support-vector-based correlation filters, respectively. Meanwhile, a database containing the keypoints of the target and the background is established and updated in real-time. After the validation of a tracking failure, the global keypoints are classified by utilizing the keypoint matching method and the target keypoints are analyzed according to these classified results. Thus the redetection results are obtained. The experimental results show that the proposed method has a better accuracy and robustness than those by the existing methods in the complex tracking scenes of motion blur, deformation, object occlusion, disappearance and so on.

An algorithm for point cloud registration in multidirectional affine transformation with variance compensation is proposed based on the statistical characteristics and shape features of point clouds, in which the problem for solving the unknown scaling factors is transformed into the problem for solving matrix eigenvalues by the overdetermined nonlinear equations, and the least square method is used for the unbiased estimation of noise variance by the quadric surface fitting. The similarity of the global vector features of point clouds is introduced, and the true value of the scaling factor is calculated by maximizing the similarity. The point cloud registration in multi-directional affine transformation is transformed into the rigid registration, and then the point cloud is registered with the main direction method. The simulation results show that the proposed algorithm has higher accuracy and smaller time consumption compared with the other existing registration algorithms when the point cloud is randomly lost or registered with noise.

In the existing structured-light vision measurement systems, the camera and the projector are installed inside a glass tube to cover the omnidirectional field of view, inevitably causing the refraction distortion. To address the distortion mentioned above, the refraction on two cylindrical interfaces is analyzed and the distortion caused by two non-coplanar refraction planes is investigated. A general distortion correction model of omnidirectional structured-light vision measurement is established. An aluminum tube with an internal diameter of 288.50 mm is measured by the omnidirectional structured-light vision measurement system and the internal diameter of this aluminum tube is calculated based on the proposed general distortion correction model. The experimental results indicate that the repeatability and precision reach 0.047 mm and 0.23 mm, respectively. The measurement system is stable and reliable, and the proposed general distortion correction model of omnidirectional structured-light vision measurement is highly accurate.

A self-supervised stereo matching algorithm is proposed based on common view. In this algorithm, the common visible region of the binocular images is determined according to the left-right consistency of disparity and thus the noise generated in the occluded region is suppressed, which provides more accurate feedback signals for the network model learning. The research results show that the prediction error of the proposed algorithm can be reduced by 11%-42% without any label data, and the performance of the proposed algorithm is comparable to that of the supervised stereo matching algorithm.

A new observation system of special vehicle based on perspective display is proposed. The hardware scale-invariant feature transform (SIFT) algorithm is achieved based on the Zynq SoC processor, and the video images corresponding to perspective are spliced in real time. At the same time, a fast electronic image stabilization algorithm based on image features is proposed for the fluctuation of driver's perspective. The experimental results show that the perspective decoding error is less than 1°, the electronic image stabilization algorithm can reduce the image shaking effectively, and the hardware SIFT algorithm can speed up the processing time of image feature calculation to the order of millisecond.

A series of phosphors such as Ca2GdZr2Al3O12…Mn4+ are synthesized by high-temperature solid-state reaction, and their phase structures and luminescence properties are characterized with X-ray powder diffractometer (XRD), fluorescence spectrophotometer and UV-visible spectrophotometer. The matrix structure indicates that Zr4+ in the [ZrO6] octahedron can be substituted by Mn4+. The XRD patterns and the luminescence intensity of phosphors synthesized at different temperatures indicate that 1500 °C is the suitable synthesis temperature. When the doping concentration (mole fraction) of Mn 4+ is 0.0050, the luminescence intensity is the strongest. When the detection wavelength is 703 nm, the excitation wavelength shifts from 343 nm to 374 nm as the increase of the Mn4+ doping concentration. Using the spectral data to calculate the crystal field parameters Dq and Racah parameters (B and C), the results indicate that Mn 4+ is in a strong field. At the same time, codoping Bi3+and Mn4+ can enhance the luminescence of Mn4+. The fluorescence lifetime test shows that the fluorescence lifetimes of codoped phosphors is longer than that of single-doped Mn4+, and there is energy transfer of Bi3+→Mn4+.

Three-cube-based fluorescence resonance energy transfer (E-FRET) microscopy is the most popular live-cell quantitative FRET imaging technique owing to its high sensitivity, no damage and fast measurement speed. To realize live-cell online real-time quantitative FRET imaging, we propose an automatic cell imaging background recognition and threshold setting method that counts gray values of an image pixel by pixel and assign the first peak gray value in the corresponding gray value-count plot as the background. The β (the empirical constant) times of the background value are set as a threshold. The corrected donor-excitation and donor-detection, and acceptor-excitation and acceptor-detection images obtained by subtracting the corresponding threshold from the raw images are used to create a Boolean logic template for data filtering of the FRET efficiency and relative concentration ratio between the acceptor and the donor via logical and operation. The results obtained through online dynamic quantitative E-FRET images of live cells expressing different plasmids on our self-assembled automatic E-FRET microscope are consistent with the expected values.

The multi-angle total internal reflection fluorescence microscopy tomography imaging technology is one of the main techniques for achieving the axial super resolution. The key algorithm is to solve the inverse problem based on the alternating direction multiplier algorithm. In order to further improve the iterative speed and convergence of the alternating direction multiplier algorithm, we propose an improved alternating direction multiplier algorithm based on relaxation factors, which is used for the solution of the inverse problems and whose core idea is to solve the relaxation process of the Lagrangian function decomposition iterative process. Based on the proposed algorithm, the multi-angle total internal reflection fluorescence microscopy imaging system is built. The image stacks with different penetration depths corresponding to different illumination angles are acquired. Then the depth information of cell microtubules is reconstructed using the improved algorithm, and the axial resolution of the system is also given. Moreover, the convergence speed comparison of the improved algorithm with the traditional alternating direction multiplier algorithm is accomplished and the range of relaxation factors for the improved algorithm to achieve the optimal convergence is also given. Finally, the three-dimensional information of mitochondrial samples is reconstructed by long time photographing, and the consecutive processes of fusion and fission are observed. The experimental results show that the improved alternating direction multiplier algorithm can be used to achieve an axial resolution of 40 nm. Moreover, the convergence speed of the iterative process is improved by more than 20% when the image reconstruction quality is ensured.

The nonlinear Schr dinger equation, which describes the propagation of electromagnetic waves in metamaterials (MMs), are studied by the Hirota method, and two exact spatially-chirped dark solitons, called Dark Soliton I (DSI) and Dark Soliton II (DSII), are obtained under the balance of linear gain and nonlinear absorption. Based on the Drude model, the formation conditions, the existence regions and the transmission characteristics of DSI and DSII in the nonlinear MMs with different electric and/or magnetic polarizations are investigated, and the analytical results are numerically verified as well. It is found that the speed of DSI existing in the negative refraction region gradually increases while the speed of DSII existing in the positive refraction region gradually decreases during the transmission. DSI has a positive chirp, while DSII has a negative chirp. In addition, the beam width of DSII is much smaller than that of DSI. Their background amplitudes and full widths at half maximum are determined by the model parameters, related to the incident wave frequency. Moreover, the transmission characteristics of DSI and DSII are quite different when the MMs are with different nonlinear polarizations. The obtained results indicate that the transmission properties of the chirped dark solitons can be controlled by selecting different nonlinear MMs or different incident wave frequencies.

To meet the needs of metal processing with different materials and different thicknesses, we design a variable spot and zoom laser cutting optical system which consists of a composite collimating lens group, a zoom lens group, a compensated lens group, and a focusing lens group. We adopt an adjustable magnification expansion beam system in the middle of the collimating lens group and the focusing lens group of the ordinary optical focusing system, and then combine the collimating lens group and the adjustable magnification beam expanding system into a composite collimating lens group to simplify the system. We establish a physical model of the variable spot and zoom optical system with four lenses, derive the geometric locus of zoom group and compensated group. Theory verification is performed with MATLAB software. Based on the existing cutting optical system, we design a variable spot and zoom laser cutting optical system, and optimize the aberrations by using software. The results show that when we move the zoom lens group and the compensated lens group, the magnification is from 1.000 to 3.750, and the focus up and down adjustable range is from -20 mm to +10 mm. The precise adjustment of the focal spot size and focus position is thus realized.

Based on the concept of supersymmetry, three supersymmetric waveguides are designed on the basis of supersymmetric optical waveguide pairs. To solve the directional mode coupling problem of supersymmetric waveguides, we insert a slant waveguide in the three supersymmetric waveguides as an adiabatic coupling medium. The coupling process is numerically simulated by beam propagation method (BPM). The effect of the slant waveguide's angle on coupling efficiency is discussed and the optimal coupling condition is obtained. Finally, a three-channel mode division multiplexer based on supersymmetric waveguides structure is designed, which provides a new idea for high-speed short-range optical multiplexing/demultiplexing technology.

Aiming at the shortcomings of low absorptivity and narrow working band for a single structure nano-antenna, we propose a multi-slot butterfly dipole nano-antenna by the fusion of the multi-slot structure and the butterfly dipole. The multi-slot butterfly dipole is formed from an Au nano-butterfly dipole etched by multiple slots. This structure can simultaneously realize the near-field coupling of tips, the grating coupling, and the hybrid coupling among different media. The coaction of these three couplings can effectively improve the absorptivity in a wide band. The absorption performance of this nano-antenna in a wide band is analyzed by the finite-difference time-domain method. The numerical analyses show that several absorption peaks exist in the absorption characteristic curve of this multi-slot butterfly dipole nano-antenna in the 400-1800 nm band, and the maximum and the average absorptivities are 98.4% and 84.1%, respectively. The absorption performance of the proposed nano-antenna is obviously superior to that of the butterfly dipole nano-antenna. This antenna can keep a stable absorption performance in a wide band under different polarization states and different incident angles of light.

A round-robin differential phase shift quantum key distribution (RRDPS-QKD) protocol based on heralded pair-coherent source (HPCS) and orbital angular momentum is proposed. The superposition state consisting of multiple orbital angular momentum states with different topological charges is used as the information carriers to remarkably increase the key generation rate. HPCS is used as the quantum source to decrease the ratio of vacuum and multi-photon pulses and to dramatically increase the key generation rate. Moreover, the propagation characteristics in the turbulent atmosphere channel are analyzed, and the influence of both channel attenuation and atmospheric turbulence on the protocol is considered.

To solve the low-classification accuracy problems of urban point clouds in complex environments, we propose a classification method based on multiscale adaptive features herein. Firstly, the classical geometric statistical features and point histogram features are combined; then, the combined feature set is used for classification basis. Random forest is then used to assess the importance of the features and adaptively select important feature sets. Finally, the point clouds are classified based on these multiscale adaptive features. Experimental results reveal that this method can achieve a high-accuracy classification for point clouds in urban areas. The proposed method can be applied to the classification of point cloud data with different resolutions at arbitrary scale.

In this study, a multi-channel analysis device based on localized surface plasmon resonance (LSPR) is demonstrated. This device comprises a broadband light source, a multi-channel precision alignment system, and a fiber spectrometer. The Savitzkky-Golay algorithm is used to process the original spectral data, and fitting curves are obtained. Further, the LSPR wavelength responses of spherical gold nanoparticles (AuNPs) with diameters of 5.0, 13.5, 25.5, and 41.0 nm in the surrounding mediums with different refractive indices are studied. The experimental results demonstrate that the LSPR peak wavelength is positively correlated with the particle size for the same refractive index, and the resonant wavelength is closely related to the refractive index of the surrounding medium. The sensitivities of the refractive index are 59.46 and 70.38 nm/RIU (refractive index unit, RIU) for AuNPs with diameters of 25.5 and 41.0 nm, respectively. The proposed device does not require any tediously long wavelength scanning procedures owing to the combination of a multi-channel alignment system and a fiber spectrometer, thereby providing an inexpensive and rapid optical detection system for conducting LSPR research.

The soil organic matter (SOM) content is an important index for evaluating soil fertility. Weigan-Kuqa region in Xinjiang is selected as the study area, based on the laboratory-derived SOM content and reflectance data, the pretreatment of Savitzky-Golay (S-G) smoothing and first order derivative (FD) are carried out. In order to further reduce the influence of sensitive band selection on modeling accuracy, we introduce the harmonic analysis (HA) algorithm to conduct the harmonic decomposition of all wavelengths. Seven principal components are obtained using dimensional reduction treatment of the principal component analysis (PCA). Subsequently, the SOM contents of soil samples are quantified by means of three methods: back propagation (BP) neural network, Genetic Algorithm (GA)-BP, and multiple linear regression (MLR). The accuracy of these methods is compared here. The results show that the correlation coefficient between SOM content and HA pretreated spectral data is improved effectively compared with those of FD data. The estimate accuracy of the non-linear model, BP neural network, is better than that of the linear model, MLR. In terms of non-linear models, the estimate accuracy of the GA-BP model is the best, with the optimal determining coefficient of 0.92, root mean square error of prediction set of 3.92×10-3, and the relative analysis error of 1.93. This study validates the effectiveness of the HA algorithm for the depth mining of spectral data, and the BP neural network model optimized by GA can improve estimate accuracy of SOM content, which can further provide scientific reference for the quantitative estimation of multiple soil properties.

Incoherent broad-band cavity-enhanced absorption spectroscopy (IBBCEAS) is gradually becoming one of the primary methods for measuring NO3 radical with the advantages of high selectivity, high sensitivity and high spatial resolution. However, due to the limited spectral resolution of the adopted spectrometers, it is not enough to distinguish the fine absorption structures of water vapor, which results in the non-linear absorption of water vapor and thus affects the accurate retrieval of NO3 radical concentration. A method based on interpolation for obtaining the effective cross section of water vapor absorption is introduced, which is used for the elimination of the interference of water vapor absorption on the concentration retrieval of NO3 radical in the IBBCEAS device. The water vapor spectra under different water concentrations are measured to obtain the effective water vapor absorption cross section by the interpolation method. The effective water vapor absorption cross section is used to retrieve water vapors with different water concentrations. The linear correlation coefficient between the retrieval results and the data from the commercial hygrometers is 0.99789. On this basis, the absorption spectra of NO3 radical and NO2 gas with different water vapor concentrations are measured and fitted. There is no water vapor absorption structure in the residual spectra, and the linear correlation coefficient between the retrieved water concentrations and the measurement values from the commercial hygrometers is 0.999. The detection limits of NO3 radical and NO2 within an integration time of 30 s are 5.8×10-12 and 3.6×10-9, respectively. This system is applied to measure the concentrations of NO3 radical and NO2 in the atmosphere at night, the measured volume fraction of NO3 radical is 18.4×10-12-22.9×10-12 with an average volume fraction of 20.2×10-12, while the measured volume fraction of NO2 is 0.6×10-9 -16.0×10-9 with an average volume fraction of 9.9×10-9. The experimental results show that the effective water vapor absorption cross section obtained by the interpolation method can be used to effectively eliminate the effect of water vapor absorption on the retrieval of NO3 absorption and to improve the accuracy of NO3 radical and NO2 gas concentration measurement.

The mass concentration detection of trace copper ion in zinc solution is difficult because of trace copper spectral signals masking, serious interference and significant nonlinearity difference of copper ion in the high and low mass concentration intervals. Aiming at this issue, we propose a spectrophotometric detecting method of trace copper ion in zinc solution based on partition modeling. The derivative spectrum combined with wavelet denoising is used to preprocess the spectral signal and reproduce the spectral peak of the copper ion to be measured. The wavelength variables are ranked by the correlation coefficient-stability value, which serves as the evaluation index of the variables, and the support vector regression (SVR) model is used to select the optimal wavelength variables. On this basis, the mass concentration of the copper ions is divided into several intervals according to the nonlinear characteristics in the mixed solution. The particle swarm optimization support vector regression (PSO-SVR) model is respectively established for each interval to compute the concentration of copper ions. The proposed method is compared with many existing regression methods. The results show that the predicted root mean square error obtained with the proposed method is reduced to 0.0678, and the model determination coefficient is increased to 99.61%. The maximum relative error obtained with the method is 6.94% and the average relative error is 2.74%.

SiO2 with gradient refractive index is fabricated via inductively-coupled plasma enhanced chemical vapor phase deposition (ICP-PECVD) technology at room temperature, and the relation between fabrication technology and optical characteristics of SiO2 material is studied. The antireflective characteristic of perovskite cells encapsulated by SiO2 with gradient refractive index is simulated by utilizing the results of ellipsometry analysis. It is found that the perovskite cells show an ultra-low reflectance, which reaches 0.5% at 550 nm. This research provides an alternative way to realize low-temperature encapsulation of perovskite cells with considering optical properties.