In order to realize imaging detection requirements of the far ultraviolet (FUV) band (140-180 nm) radiation of N2 LBH in the auroral spectrum, we used the far ultraviolet photon counting imaging detector to detect it. The detector is mainly composed of CsI photocathode, V stack microchannel plates, induction type wedge and strip anode. First, the calibration device of FUV radiation is constructed, which is mainly composed of vacuum tank, vacuum displacement platform, standard transfer detector, FUV grazing incidence monochromator, vacuum ultraviolet light source, inspection equipment of signal processing, etc. Then, the quantum efficiency, resolution, dark noise, limit count rate, and other main performances of the detector are tested in detail. Finally, test data obtained are analyzed. Test results indicate that the maximum quantum efficiency of the detector at the working waveband is 12.9%, spatial resolution is 88.3 μm, dark count rate is 0.87 counts/(s·cm2), and the effective test range of the count rate of the detector system is 0-350000 counts/s. The technical specifications of the developed detector meet the requirements of radiation imaging for N2 LBH in the auroral spectrum.

Aiming at the requirements of oceanic profile observation and problems of traditional expendable bathythermograph(XBT), we propose a kind of optical fiber temperature-depth oceanic profile sensor based on fiber Bragg grating-long period fiber grating (FBG-LPG) cascaded structure (fiber XBT), and design and optimize the optical path by using the reflective optical fiber end mirror. The experimental results show that, when the end reflectivity of fiber is in the range from 48.8% to 72.5%, the resonant peaks of FBG and LPG are reasonable and the whole spectrum is suitable for demodulation.

Based on the atmospheric weak turbulence channel model, a constellation recognition method based on fuzzy clustering and improved back propagation (BP) neural network is designed. The fuzzy C mean (FCM) algorithm is used to get the cluster center of the wireless optical multiple phase shift keying (MPSK) subcarrier signals constellation. By calculating the hard tendency of the fuzzy classification, we obtain the feature of constellation. Finally, improved BP neural network as the classifier is designed and used to accomplish the modulation recognition. When the log-amplitude fluctuation variance σχ2=0.1, the correct recognition rates of four different modulation styles are all up to 100%. With the increase of fluctuation variance, convergence of MPSK signal constellation diagram becomes worse, but the total correct recognition rate also gets to 87.5%, and the recognition rate of 16 phase shift keying (16PSK) is improved obviously.

Based on the principle of inter mode interference, we apply the single mode-multimode-single mode (SMS) optical fiber structure to the optical fiber security system. In the system, multimode optical fiber is used as sensing fiber, and wavelet transform is used to realize the detection of external vibration signals. The theory of intermodal interference of multimode optical fiber is analyzed. The sensing optical fiber is arranged in the fence and laid on the ground, respectively. Several vibration invasive events are analyzed in experiment. The experimental results show that the sensing system of this structure can detect the vibration invasive signal quickly and effectively, and it has good feasibility in the field of security in optical fiber. The structure sensing system maintains the advantages of simple structure, low-cost, which has very important significance in optical fiber security systems.

We propose a novel fiber current sensor based on single light path detection and fiber loop scheme in order to solve the problem of low current sensitivity of traditional single light path detection fiber current sensors. Fiber loop scheme of this fiber current sensor is built by single-polarization single-mode couplers, which replaces the polarizer and the analyzer in traditional single light path detection fiber current sensors. This design not only keeps the advantages of simple structure and low cost, but also improves the current sensitivity of fiber current sensor via repeated cycle amplification, which inherits from the loop scheme. Both theory and experiment show that the system current sensitivity increases with cycle number K and the new fiber current sensor is feasible.

Regenerated fiber Bragg grating (RFBG) cannot be directly used in practical engineering for temperature measurement due to its low reflectivity, so a high-temperature fiber laser sensing method based on low-reflectivity RFBG is proposed. RFBG is used as one mirror of the resonant cavity, and a length of unpumped Er-doped fiber(EDF) as saturated absorber is adopted to compress line width and suppress multi-longitudinal mode of laser. The current threshold of fiber laser is 68.9 mA. At 150 mA, the output laser is stable and has a good linear relationship with temperature when temperature varies in the range of 300 to 800 ℃. In the temperature rising and falling test, the correlation coefficient is up to 0.99974, and the average temperature sensitivity is 15.41 pm/℃. During 3 h of laser stability test at 700 ℃, the maximum variation of the laser wavelength and intensity is 0.032 nm and 0.409 dB, respectively. The results show that the signal-to-noise ratio is higher than 50 dB, the output laser is stable, and there is no mode hopping phenomenon in the temperature-rising and temperature-falling process.

A high-speed optical sampling technique based on optoelectronic oscillators is proposed. Optoelectronic oscillators generate sampling pulses with high repetition rates and low time jitter directly, as a result, the time-domain multiplexing of low repetition rate sampling pulses is avoided,the system complexity is reduced, and the system efficiency is improved. The generated pulses are used to sample signals by time-domain Fourier transform, the amplitude information of the signals at different moments are modulated to the detection light at different wavelengths,then the light with different central wavelengths is filtered by wavelength division multiplexer or tunable filters, thus high-speed real-time sampling and serial-parallel conversion are realized, and electronic analog-to-digital conversion quantization with low rate is well matched. The optical pulses with repetition rate of 10 GHz and time jitter of 195 fs are generated in the experiments. Three kinds of signals with different waveforms are sampled, the sampling waveform is obtained, and the optical sampling rate of 40 GSa/s is reached. The experimental results show the feasibility of the proposed optical sampling technique based on optoelectronic oscillators.

To convert Gaussian spot into flat line spot, a spheric-aspheric cylindrical lens is designed. The design is accomplished by using Zemax programming language to set operands and default optimization function. The design is compared with the aspheric lens-cylindrical lens group with the same performance parameters. The influence of the distance from the last surface to the image surface of the spherical-aspherical cylindrical lens on the length-width ratio of the linear spot and the flatness of the line-focused is analyzed. Compared with the aspheric lens-cylindrical lens group, the line spot size is the same but the flatness is normal (about 10% decline at the edge). Adjusting the distance from the last surface to the image surface in a certain range can make the flatness better (can reach 90%), but the aspect ratio is reduced (from 88.33 to 20.38). The results show that, in order to obtain the line spot with uniform distribution of energy, in the adjustable range of length-width ratio, the design of spheric-aspheric cylindrical lens can simplify the structure of beam shaping system and meet the requirements of lightweight. It is a feasible method.

The non-uniform illumination and color deviation lead to the difficulty in haze removal for nighttime image. The current image dehazing methods are mostly designed for daytime images. There are few studies on nighttime image dehazing. Therefore, we propose a new nighttime image dehazing method based on the structure-texture image decomposition model. Firstly, the haze image is divided into a structure layer and a texture layer. Secondly, to estimate and then optimize the initial atmospheric light, the median filter and the weighted norm L1 regularization model are introduced in the structure layer. After that, dehazing and color correction are performed. Thirdly, the transmittance is estimated with discrete cosine transform coefficients in the texture layer. Finally, the ultimate haze-free image is recomposed with the texture layer and the haze-free structure layer. The experimental results show that the proposed algorithm is effective in the nighttime haze image processing, generating haze-free images with clear details and natural colors.

Intelligent robot has recently matured in industry, whose core technology is machine vision, especially object recognition. In existing object recognition methods, scenes are segmented based on color, and features are then extracted to recognize objects. However, over segmentation exists for scenes with complex color features, which influences subsequent object recognition process. To deal with the over segmentation problem, a multi-object image segmentation algorithm based on local features is proposed, which uses binocular camera to collect scene images. Firstly, the scene image is preprocessed. The depth information of the scene is then obtained by stereo matching, and is used to determine the target area. Secondly, the local features of the target region are extracted by a scale-invariant feature transform (SIFT) algorithm with dynamic threshold, and the local features are then transformed into feature constraints. Finally, the feature vectors, which are based on region constraint, feature constraint and spatial information, are used for clustering segmentation to obtain the final segmentation result. Simultaneously, each object region is recognized. The experiment results show that the overall error rate of the proposed algorithm is less than 10% for a scene with complex color features, and is reduced by 15% compared with those of the existing algorithms.

A fast global registration algorithm for 3D point cloud is proposed, which is used to estimate the rigid pose relationship of any two sets of similar point clouds in space. First, the three principal component vectors of two groups of point clouds are calculated to form their own principal component (PC) coordinate systems with their respective center points. Then, in order to obtain the approximate pose relationship between the two similar groups of point clouds, the coordinate transformations of the two groups of point clouds are respectively performed, and the directions corresponding to the PC coordinate axes are corrected by the mean of Euclidean distances of some close points. After the above coarse registration, the two groups of point clouds can be matched fast and accurately at arbitrary position by the fast iterative closest point (ICP) algorithm. The experimental results show that the proposed method can achieve global registration for any two sets of point clouds with similar shape and completeness at any position in any pose, and has higher speed and accuracy.

Aiming at the problems of the insufficient sampling and sensitivity to random noise and non-uniform illumination of the local binary pattern, a face recognition method of the improved gradient local binary pattern (IGLBP) is proposed. Two groups of 3 pixel×3 pixel subneighborhood are obtained by the multi-radius and multi-direction sampling mode, including 16 pixels in two radii and eight directions. The features are extracted by the gradient local binary pattern, and then the two sets of them are encoded to produce IGLBP. Finally, the IGLBP feature is divided to get the feature vector of the face according to the block histogram, and it is used for classification and recognition. The experimental results of CAS-PEAL and AR face database show that the proposed algorithm can effectively extract the feature information, and it is robust to variations of the illumination, expression, partial occlusion and noise in face recognition.

The stereoscopic image distortion can affect the edge, structure, depth and other information of image. In this paper, we propose a no-reference stereoscopic image quality assessment metric based on the human eyes'comprehension of image's low-level structure. First, the left and right views, cyclopean map and disparity map are decomposed by the dual-tree complex wavelet transform. Second, the phase amplitude characteristics of the wavelet sub-band of the left and right views, cyclopean image and disparity map are extracted. Similarly the gradient features of the wavelet sub-band of the left and right views and cyclopean image are extracted. Finally, these features are feeded into the support vector regression to train the mapping model for predicting the quality score of tested stereoscopic image. The experimental results on LIVE3 DIQD Phase 1 and LIVE3 DIQD Phase 2 show that the proposed method is highly correlated with the human visual system, achieving excellent prediction performance.

For detection of the end face defects of small cylinder core with complex image background, the current detection methods based on threshold, morphological and edge Canny are inaccurate due to the interference of the gray information caused by the unevenness of the surface coating. We present a defect detection algorithm based on machine vision for optical triangulation. Firstly, two images with certain defect information are obtained based on the active light visual inspection method. Then, the defect contours are sketched out according to the template matching, morphological analysis and frequency statistics. Finally, we use Delaunay triangulation technique to realize the accurate sketch of the defect area. The experimental results show that this method has high accuracy and robustness for defect detection hidden in complex background, and the repeatability can reach sub-pixel level, with good anti-jamming, strong practicability and high efficiency.

In order to overcome the shortcomings of the traditional binary image boundary tracking, including slow speed and leakage tracking, an algorithm of binary image boundary tracking and boundary chain code acquisition with strong generality is proposed based on the Freeman chain code. Based on the improvement of traditional Freeman chain code search algorithm, the average search direction is reduced to 2.5 when a new boundary point is found. Further the complex boundary segment is processed, and the starting point of the previous boundary chain is used as the starting point for the following boundary chain search, and fills the following boundary points, so that the image scanning from top to down and left to right is only one time to track all the boundaries, and eliminates the single pixel redundancy by setting the thresholds. The experimental results show that the proposed algorithm can effectively track the complex mesh boundary and internal holes, and it has the ability to remove the redundant points with anti-noise and faster speed.

Segmentation evaluation is an important way to improve the performance of algorithms. A gray evaluation model is proposed based on combinational weighting, aiming at the problem that the current index of image segmentation can not reflect the results of segmentation well. Firstly, variation of information, global consistency error, and probabilistic rand index are selected to evaluate the quality of image segmentation. Secondly, a subjective and objective combinational weighting method is proposed which combines Delphi method, forced decision method, and entropy method. The weights not only reflect the subjective preferences of observers, but also highlight the objective differences of images. Finally, the proposed model is used to make a comprehensive evaluation of test images. Experimental results show that the proposed evaluation model is consistent with the subjective evaluation results and the real ground results. Moreover, this model is used to compare the segmentation results of the maximum entropy threshold algorithms based on flower pollination algorithm, genetic algorithm, and shuffled frog leaping algorithm, respectively. The obtained rank is consistent with the result of maximum entropy, which further validates the effectiveness of this model.

In order to solve the problem of inter-class difference caused by the angle of view and scale change, we propose a method based on histogram of spatio-temporal oriented principal components of three-dimensional (3D) point clouds for action recognition. Firstly, the depth sequences are converted into 3D point clouds sequences. Then, we use a novel image preprocessing method to get new depth sequences. Namely, the sampled depth sequences are limited in spatio-temporal dimension to remove areas with less information, and reduce the redundancy of the input data and the influence of space scale change In order to solve the problem of weak correlation between frames, we adopt histogram of spatio-temporal oriented principal components (HSTOPC) method to describe 3D point clouds sequences and obtain the direction of each point of the 3D point clouds in sequences. For all direction of 3D point clouds in sequences, we use multilayer overlapping segmentation method to obtain HSTOPC descriptor. Finally, we use the support vector machine classifier for training and test. Experimental results on three human action recognition datasets show that the proposed HSTOPC feature descriptor has better robust for noise, rate variations, view change and temporal misalignment, and is able to improve the accuracy of human behavior recognition significantly.

In view of lacking of the texture information and the edge information in the infrared image, the accuracy of depth estimation is hard to be improved. We propose a deep neural network to estimate the depth of infrared images. The network combines a two-dimensional (2D) residual neural network and a three-dimensional(3D) convolution neural network. The traditional methods of estimating the depth of a single infrared image omits the interframe information and is prone to fuzzy or even missing object contour. The 2D and 3D network inputs are added dense optical flow and the frame before and after the image, respectively. Secondly, the feature map extracted from the 3D convolutional network is further connected to the feature maps of the 2D residual network. Unlike the fully connected layer of the traditional neural network, fully convolutional layer breaks through the size constraints of the input. The experimental results show that the accuracy of the proposed infrared image depth estimation method is improved, and the object contour estimated is clear and complete.

In order to solve the problems of traditional object classification methods, such as the inaccurate expression of spatial structure features, and the classification model parameters trapped in local optimum, we propose a method that combines the overlapping pyramid method with the Bayesian optimized neural network. Firstly, we extract the contour fragments of different lenghts from the object contour as features, and encode them with the locality-constrained linear coding encoder. Then, the proposed spatial overlapping pyramid histogram is used to represent the images. Finally, the Bayesian optimized feedforward neural network classifier is used to accomplish the classification. The experimental results based on the standard Animal dataset show that the accuracy of the proposed method is improved by 1.4% as compared to the Bag of Contour Fragment method, indicating that the proposed method can accurately represent the context and structure of the shape and is effective in object classification.

Based on the principle of sample block texture transmission, the influences of the source texture image’s texture information and structural information, and the structural information of the target image on the stylistic effect of the texture transfer are studied. The source texture image and target image are decomposed with the relative total variation model of non-filtering for eliminating the structure information of the source texture image and the texture information of the target image. Texture transmission algorithm is used for the texture transmission of the above reserved information image. The improved algorithm avoids covering the target image structure when the structural information of the source texture image transmits in the traditional algorithm. In this way, the edge structure information of the target image and the transmission result image are superimposed, which enhances the edge information of the transmission result graph, and improves the stylized effect. Experimental results show that the improved algorithm can achieve better transmission stylized effect than that of the traditional algorithm.

In order to solve the problem that more redundant pixels and less consideration of picture quality of the region of interest (ROI) when the 360° panoramic video is encoded in an equirectangualr projection plane. A new ROI-based 360° panoramic video coding method is proposed, which uses the current coding residual frame information to get and guide the next frame coding. The pixel redundancy at each latitude is calculated by the mapping function from the sphere to the projection plane, which is used as the adjustment factor of quantization parameters of non-ROI. The quantization parameters of ROI and non-ROI are utilized to determine each frame's encoding parameters in largest coding unit level. The experimental results show that, compared with the HEVC encoder, for the same weighted spherical peak signal-to-noise ratio and spherical peak signal-to-noise ratio, the proposed approach can achieve an average rate reduction of about 2.46% and the most of 4.98%, for the same coding rate, it can achieve an average video quality increased 0.145 dB. The content of ROI improves obviously compared with the related representative methods.

A novel spectral imaging method based on multi-modes in time-frequency domain is proposed. This method is used for the nondestructive detection of four kinds of ceramic matrix composite (CMC) samples with defects. A THz image set for the test samples is formed, and five quantity evaluation indexes for image quality are introduced. The THz images with the best quality are chosen via the index integration treatment. The retrieval of this image is realized based on the Scale Invariant Feature Transform (SIFT) feature selection and K-means clustering algorithm. The experimental results confirm that the proposed method can be effectively used in the imaging detection of the CMC materials with various defect widths at different positions.

Since the digital mircomirror device (DMD) is working in slant optical path in the coded aperture imaging spectrometer, the coded pixel will generate unsymmetrical distortion on detector, which will make it unable to determine the coded method corresponding to each pixel of obtained coded image when decoding. In order to solve this problem, this paper proposes a method of "unsymmetrical deformation regular stripe calibration". Through the changes that the regular stripe produces on the detector, the deformation of the coded pixel is visually observed. According to the known stripe rule, the deformation of the image can be quantitatively analyzed and corrected. This method can ensure the correction of the coded image under the premise of clear imaging of the entire field of view of the system. Firstly, the imaging principle of the designed spectrometer and the deformation reason of the coded pixel are introduced. Secondly, the detector is adjusted during the experiment to obtain a full-clear field of view. Finally, the proposed method is used to process the coded image. Experiments show that the similarity between the processed image and the theoretical value is 37.87% higher than that of the unprocessed image. The image restores the pattern shape loaded by the DMD, which lays the foundation for the next decoded operation.

In the structured light 3D measurement system using the commercial projector as the source, all the encoding patterns are projected with the two-dimensional matrix form. The computer transmits the patterns to the projector, and the patterns are projected onto the scanned object by two-dimensional images. If the code is a striped pattern, the characteristics of stripe structured light coding that the gray information remains unchanged in the same line or in the same column are not considered. Therefore, according to the unidirectional characteristic of the fringe encoding, we present a design scheme of a special structure projector, which can automatically extend one-dimensional gray sequence into encoded structured light. One-dimensional coding sequence with all gray information and a designed pixel-extended circuit are used to expand the pixel of structured light coded image, in order to improve the projection efficiency of structured light 3D imaging. The experimental results show that the speed of projection is improved by about 71.8% compared with liquid crystal display projection method, and the order of magnitude of phase accuracy difference is 10-3. The projection efficiency of dual encoded structured-light images in 3D measurement is effectively improved.

A multi-point measurement method for illuminance uniformity of lithography machines is proposed. The method of using ultraviolet (UV) enhanced PIN photodiode can rapidly measure the light intensity of the exposure machine, and the combination of improvement on the basis of traditional current-voltage amplifying circuit and composite amplification greatly improves the repeatability of exposure machine illuminance uniformity multi-point measurement method, and makes every single point measurement repeatability of measurement below 0.02 mW/cm2. Experimental results show that with the use of UV enhanced PIN photodiode exposure machine light source, multi-point measurement consistency is less than 0.1 mW/cm2, and the detection of the illuminance uniformity meets the requirements.

Phase measuring deflectometry (PMD) based on fringe reflection has been widely studied as a way of obtaining three-dimensional shape of specular objects. System calibration is an important step, and it determines the accuracy of the measurement results. We propose a calibration method to obtain the system parameters based on phase information. As a result, it can build the relationship between the absolute phase map and depth data. A contrast experiment is done for verification about extrinsic parameters of the LCD screen by phase data and the checkerboard. The experiment shows that the method using phase data is more accurate when images are out of focus. Using the calibration system, we test a concave mirror and an artificial specular step with discontinuous reflective surface, and the error is about 22 μm. Experiment results show that the proposed method can precisely determine the system parameters, so that 3D shape of specular objects can be measured with a high accuracy.

We introduce the principles of conventional temperature measurement method and three-band infrared thermal imaging temperature measurement method for sea surface, then solve sea surface temperature with these two methods, separately, and comparatively analyze the measuring accuracies of the two methods under different observation directions and infrared imager precisions. The results show that three-band infrared thermal imaging temperature measurement method for sea surface can eliminate the influence of the sea surface emissivity error on the temperature measurement accuracy. When the infrared imager precision is within ±0.25 K, the deviation of sea surface measured temperature is less than 0.5 K, basically, so its accuracy is relatively high. Improving the infrared imager precision can effectively improve the accuracy of three-band infrared thermal imaging temperature measurement method for sea surface. When the zenith angle of observation direction is in the range from 70° to 85°, the deviation of conventional temperature measurement method is much larger than that of three-band infrared thermal imaging temperature measurement method for sea surface, so we must use three-band infrared imaging temperature measurement method for sea surface in this case.

A novel tunable laser based on Fabry-Perot (F-P) laser self injection locking is proposed. A single longitudinal mode is filtered from the multi-longitudinal modes of the F-P laser by the F-P etalon and injected back into the F-P laser to suppress side mode. We can change the spectrum of the F-P laser by tuning the driving current rapidly, and realize a wide tuning range of the wavelength by virtue of the vernier effect between the varying spectrum of the F-P laser and the stable spectrum of the F-P etalon. The static characteristics of the tunable laser are studied and 18 sets of single longitudinal modes are available. The tuning range of wavelength is from 1535.2 nm to 1548.68 nm, and the side mode suppression ratio of the single longitudinal mode is 40 dB. The power difference of each wavelength is less than 15 dBm and the frequency interval is 100 GHz. A wavelength tuning experiment is carried out, and the tuning time for different wavelengths ranges from 60 ns to 100 ns. The feasibility of the scheme is verified.

The surface treatment of TC11 titanium alloys by lasers with different energy spatial distributions is conducted, and the residual stress in specimens induced by laser shock peening (LSP) is tested by X-ray diffraction. The results indicate that, under the same process parameters, the surface residual stress and the stress-affected layer depth in titanium alloys after LSP with a Gaussian distribution are -636.71 MPa and 1187 μm, respectively, while those with a flat-top distribution are -559.20 MPa and 811 μm, respectively. The theoretical calculation shows that the peak stress of shock waves induced by Gaussion spatial energy distributions is 1.55 times of that by flat-top spatial energy distributions.

The 6061 aluminum alloys are treated by laser shock peening (LSP), and the effects of LSP on their wear resistance and electrochemical properties are investigated. The micro-hardness, residual stress, and so on are examined and analyzed. The results show that, LSP can effectively enhance the mechanical properties of samples. On the surfaces of the LSP treated specimens, there is no phase change, but the micro-hardness and residual stress increase obviously, and the corrosion resistance of 6061 aluminum alloys is obviously improved. The friction coefficients before and after the LSP treatment are comparable, but the wear loss of the LSP treated specimen decreases and the wear resistance is significantly improved.

The process parameters of laser bending forming of copper-nickel composite sheets are screened based on the experimental design of parameters. The central composite design of experiments is chosen for the experimental plan and the response surface methodology is adopted for the establishment of mathematical model among all process parameters. The interactive influence law of each process parameter is disclosed. The satisfaction function is used to optimize the process parameters and the optimal combination of parameters is obtained.

The interference intensity distribution of four-beam interference with right-hand circular polarization is theoretically studied and simulated with MATLAB. The impact of incident angle on the intensity distribution of interference is studied. The results show that there is a uniform two-dimensional periodic intensity distribution when four beams show symmetrical distribution. The peak intensity appears at the intersection of three lines with the slopes of S1=+1, S2=￥, S3=-1, respectively. And there are certain periods dx=dy=λ/sin θ in the x and y directions. When incidence angle θ changes, the period of the pattern changes accordingly but not the symmetricity of pattern. However, when the incident angle or azimuthal angle (θ1,α1) changes, both the slopes Sl(l=1,2) and periods dxl,dyl(l=1,2,3) change.

For the space applications of high-power fiber lasers, we experimentally investigate the radiation performance of fiber lasers in radiation environment. The two fiber lasers, pumped at different wavelengths of 976 nm and 915 nm , are used with a same kind of 20/400 μm Yb-doped gain fiber in the experiment. After 12.4-krad irradiation by the γ ray produced by Co 60, the laser output powers of the 976-nm-pumped and 915-nm-pumped fiber lasers decay from 32.68 W and 32.04 W to 20.09 W and 5.63 W, respectively. The results show that the 976-nm-pumped fiber lasers are more radiation-resistant than the 915-nm-pumped fiber lasers. In addition, passive fiber devices, such as fiber combiners and cladding strippers, are used for irradiation experiments. After the irradiation, the coupling efficiency of fiber combiner is reduced, and the stripping degree of cladding stripper is increased. It is shown that irradiation has an effect on passive fiber devices.

304 stainless steel joints with 5 mm thick are welded by laser-MIG hybrid welding with 10 kW disc laser. The influence of the laser-arc distance on the forming and mechanical properties of the welding seam is investigated, and the influence mechanism of the arc on microstructure and microhardness of the welding seam is analyzed. The results show that an optimized laser-arc distance not only have a synergistic effect between the laser heat source and the arc heat source, but also increases the stability of the droplet transfer. When the laser-arc distance is 1 mm, which makes an optimized distance, the droplet transition is stable and the welding seam is formed well. The tensile strength is 631.33 MPa, which is reaching to 98.6% of the strength of base metal. The microstructure of the welding seam is composed of austenite and skeletal and lath ferrites, and the microstructure of the upper part (Ⅰpart) of the welding seam is smaller than that of the middle part (Ⅱpart) and lower part (Ⅲ part) of the welding seam. The microhardness for the upper part of the welding seam is about 175 HV, which is equivalent to that of the base metal, the microhardness for the middle part of the welding seam is about 180 HV, and the microhardness for the lower part of the welding seam is about 190 HV, which are all lower than that of the heat affected zone.

In order to improve the passband flatness and 3 dB modulation bandwidth of the distributed feedback (DFB) lasers, we study the relationship between the device parameters (such as grating coupling coefficient, cavity length and electrode ratio) and the frequency response of 1st and 2nd order grating push-pull modulated DFB lasers by using one-dimensional traveling wave model. The parameters used in the simulated model are extracted from the photoelectric characteristics of the actual devices. The simulated results show that, for the DFB lasers with 1st order grating structure, the push-pull modulation bandwidth is 12 GHz larger than the single-electrode direct modulation bandwidth at the appropriate grating coupling coefficient, cavity length and electrode ratio. For the DFB lasers with 2nd order grating structure, the push-pull modulation bandwidth is 40 GHz larger than the single-electrode direct modulation bandwidth. Besides, the passband flatness of the 1st order and 2nd order grating push-pull modulated lasers is much better than that of the ordinary direct modulation lasers. In addition, the results of the experiments show a substantial increase of the bandwidth with push-pull modulation, even for general design of active region.

Pantograph is an important equipment of current collector for the electric train. The abrasion of the pantographic slide affects the coordination of pantograph-catenary system. Automatic measurement of pantographic slide abrasion is one of the important means to realize the automatic fault inspection of pantograph. According to the image characteristics of pantographic slide captured by the vision measuring system, an accurate recognition algorithm for upper and lower edges of the slide is proposed. The proposed algorithm is fulfilled by employing the accurate sub-pixel edge location method based on partial area effect. By comparing with other common edge extraction algorithms, the feasibility and superiority of the proposed algorithm are proved.

Kernelized correlation filters (KCF) algorithm tracker updating its model parameters at every frame makes it unable to effectively deal with the problems of fast motion and interference of the target in most environments. A nuclear-related object tracking method based on enhanced threshold updating is proposed. Based on the average peak correlation energy (APCE), the APCE threshold and APCE gradient threshold are combined to determine the reliability of the tracking results, which are used to determine whether the model is updated. In this paper, the APCE threshold value is reversely enhanced and the APCE gradient threshold is positively strengthened. When the APCE and APCE gradients are all higher than the respective thresholds, the model is updated, otherwise, the model will stop updating. The quantitative and qualitative experiments show that the algorithm is more effective than the KCF algorithm for the fast motion and interference of the target. The proposed algorithm also provides a good reference value for the design of tracking algorithm based on the idea of gradient detection tracking performance and threshold enhancement.

In view of the fact that the traditional tracking learning detection (TLD) algorithm has poor robustness, low tracking success rate and low computing efficiency, a TLD tracking algorithm combining binary robust invariant scalable keypoints (BRISK) feature points and region prediction is proposed. In the tracker, the BRISK feature point is combined with the conventional pixel points used for tracking, and they are used for target tracking together. Due to the fast extraction of BRISK feature points, the total computing time of the tracker is reduced. In the detector part, the combination of Kalman filter and Markov model direction predictor greatly reduces the number of sub-image blocks sent to the detector, and enhances the identification ability for similar targets, thereby improving the speed and accuracy of the detector. The experimental results show that, compared with the traditional TLD algorithm, the tracking accuracy of the proposed TLD algorithm is improved by about 64.4%, and the running speed is increased by about 39.6%, and its robustness is better.

The influence mechanism of the internal stress in quartz on the precision of digital sun sensors is theoretically investigated. The stress birefringence test of quartz specimens is conducted and the different quartz specimens are mounted on the digital sun sensors for a full visual field test. The means to control the stress level of quartz materials are put forwards. The research results show that there exists interference spots only for the quartz specimens with locally concentrated internal stresses, which is consistent with the theoretical analysis result.

The reflectivity, internal and external quantum efficiency, minority carrier lifetime and electrical characteristics of mono-crystalline silicon solar cells with double-layer and three-layer silicon nitride antireflection coatings are compared. The optical characteristics of double-layer and three-layer silicon nitride antireflection coatings are simulated. The results show that, the three-layer silicon nitride antireflection coating has a better antireflection effect and a better output characteristic. The double-layer and three-layer silicon nitride antireflection coatings are fabricated on P-type mono-crystalline silicon by the plasma enhanced chemical vapor deposition (PECVD) process. The analysis results of their antireflection and passivation effects show that the three-layer silicon nitride antireflection coating has a smaller reflectivity and a better passivation effect compared with the double-layer silicon nitride antireflection coating. The electrical characteristic testing results of the mono-crystalline silicon solar cells with double-layer and three-layer silicon nitride antireflection coatings show that the conversion efficiency of the mono-crystalline silicon solar cells with three-layer silicon nitride antireflection coating is relatively higher than that with double-layer silicon nitride antireflection coating.

The effects of La/Ce single-doping and co-doping on the electronic structure, differential charge density, density of states and optical properties of TiO2 are investigated based on the first principle method. The results show that, because of the synergistic action among La 5d, Ce 4f, and Ti 3d electronic states, as for the TiO2 under La-Ce co-doping, the downward shift of the conduction band is larger, the band gap is smaller, the redshift of the absorption spectrum is more obvious, and the photocatalytic performance is better than the corresponding ones under single-doping.

A novel broadband absorber in microwave frequency band based on metamaterials is designed, which is a typical three-layer structure with top and bottom layers of metal and the middle layer of dielectric. The top layer is a metal structure which is designed on a single cell by using the principle of multiple-absorption-peak superposition and composed of a circular open ring and a square open ring. The simulation results show that the absorptivity of the proposed absorber is over 90% in the frequency range of 10.65-22.39 GHz, the full width at half maximum (FWHM) is 13.07 GHz, and the relative FWHM is 83.7%. The overall thickness of this proposed absorber is only about 1/10 of the center wavelength.

The traditional algorithm of voxels growth is improved to ultimately achieve the purpose of extracting the local blood vessels of the heart. The initial voxels seed spots are delineated on the local blood vessels of interest within the three-dimensional model of the heart. The initial voxels set adds new voxels, which are similar to the initial voxels set in terms of their gray-scale histogram. Newly added voxels as new seed spots continue to look for similar voxels according to growth rules. In order to prevent leakage problems during voxel growth, we sets the termination rule to limit the boundaries of voxels growth, and discover all the voxels included in the blood vessels. The improvement of growth rules and termination conditions of the voxels growth algorithm, to a certain extent, solves certain problems of the traditional voxels growth algorithm such as poor interaction and not real-time running. The algorithm has the advantages of high accuracy and good robustness, hence it is proved to be an effective imaging aid tool for the clinical treatment of cardiovascular diseases.

Pulse laser with 50 MW maximum peak power and wavelength of 531.81 nm is used to pump 500 m silica single mode fiber (SMF) in order to induce the stimulated nonlinear effect. Based on the nonlinear optics theory, optical fiber transmission theory and experimental measured data, effect of the four-wave mixing (FWM) on the spectral structure of stimulated Raman scattering (SRS) is studied. And the transmission modes of SRS and FWM in the silica SMF are also presented. Research results show that in the silica SMF, FWM effect is able to induce the low order Stokes light of SRS to produce additional peak spectral lines. Meanwhile the spectral broadening effect of the high order Stokes spectral lines of SRS can also be produced, and this could finally lead to frequency shift of the high order Stokes spectral lines of SRS.

The photorefractive index generated inside the photorefractive crystal has a very small change and is not easy observed in real time, resulting in the failure to characterize the quality of the writing lattices.In order to observe the temporal and spatial variation of the photorefractive index caused by the writing of photorefractive photonic lattice, we propose amplified Fourier transform method based on Mach-Zehnder optical path. The resolution of the measurement is greatly improved. This method can observe the time and space changes of refractive index of high density photonic lattice in real time, at the same time, it can identify the quality and uniformity of photonic lattice, and solve the time control problem of high density photonic lattice fabrication quality. Compared with the indirect method, the proposed method is more straightforward, the measured data is more accurate and clear, and the resolution can be easily determined by the combination of Fourier transform lens.

The influence of material dispersion on the output characteristics of cross-polarized wave (XPW) in BaF2 crystal driven by Ti∶sapphire laser pulses is investigated theoretically. The numerical simulation results reveal that, when the material dispersion is considered, the energy conversion efficiency of the crystal decreases owing to the effect of the dispersion. The pulse duration of XPW is longer than 1/3 duration of the incident pulse, and the saturation intensity increases. The results indicate that the influence of material dispersion on the output characteristics of XPW driven by short laser pulses cannot be neglected.

Field of view (FOV) and resolution are very important parameters for head-mounted displays (HMD). However, so-called "resolution/field-of-view invariant" prevents single-channel HMD from achieving both high resolution and wide FOV. This limitation can be effectively broken by the tiled method. A binocular optical-tiled HMD with four spherical lenses that have been cut precisely is designed. Each eye includes two channels. For an eyepiece consisting of two channels, the horizontal FOV is 111.4° and the maximum vertical FOV is 90°. Three 6 inch (15.24 cm) liquid crystal displays with resolution of 1920 pixel×1080 pixel are used in this prototype as image source. The monocular angle resolution reaches 3.85 arcmin. 3D printing is also used to make the external structure of the HMD prototype. Key issues in the tiled system, such as exit pupil and exit pupil distance, are analyzed in detail. Based on these discussions, the intensity distribution of the human eye retina is simulated in LightTools. In addition, according to the distorted data of the lens design, the distorted image after processing is obtained. The image display experiment is performed on the prototype, and the results verify the feasibility of the proposed method.

In order to realize the rapid and accurate evaluation of luminous quality of light guide plate (LGP) and find nonuniform areas to optimize, we propose a new design method based on image processing technology, the evaluation criteria of luminous quality of LGP is deeply researched. The light distribution is represented by gray histogram, the evaluation formula of uniformity is given and the concept of evaluation precision P is presented. Furthermore, the dots position corresponding to discrete grayscale is located with program, the accurate optimization is realized. The method is verified and simulated by taking 184 mm×314 mm×0.55 mm LGP for example. The relation between uniformity and precision is given by analyzing three stage irradiation maps based on gray histogram in the optimization process. The method is compared with conventional evaluation methods. Results show that the gray histogram method can evaluate luminous quality of LGP objectively and comprehensively on the premise of precision and the possible accidental deviation is avoided. The optimization speed and precision are improved. The uniformity of LGP can improve 64% after optimization. The accidental deviation is reduced gradually with TracePro and nine dots in the optimization process. What's more, the results tend to consistent with three evaluation methods.

Commercial coating design software based on optical transfer matrix theory and traditional method of correlation analysis are used to obtain a series of simulated reflection spectrum of ultraviolet AlGaN-based Bragg reflector (center wavelength of 314 nm). The correlation between partial structural parameters and reflectivity or center wavelength are studied, as well as the correlation coefficients. The results of correlation analysis show that the thickness of the AlGaN layer is the most impact factor on the reflectivity and the center wavelength, which is a negative correlation. Bases on this conclusion, we obtain an optimal UV AlGaN DBR structure with high reflectivity and wide detection band. Furthermore, the correlation analysis and optimization results match the actual data, which provide a new way and method for the design and application of UV AlGaN DBR.

The classic Cassegrain system combined with a front lens group is chosen and a high-power laser-beam focus system over a variable long distance is designed based on the primary aberration theory. When the moving distance of the target objects is within 500-5000 m, it is achievable to control the focusing spot size by the adjustment of the distance between the front lens group and the Cassegrain system and the different object distances can be reversed from the location of the lens group. The initial structure is optimized with the Zemax optical software, and the design results show that the spot sizes under different object distances all meet the design requirements.

A single remote sensing image super-resolution reconstruction method based on online variational Bayes expectation maximization coupled dictionary learning is proposed in this study to improve the spatial resolution of low resolution remote sensing images. The method first establishes the probability distribution model of the dictionary atom and each parameter, divides it into local variables and global variables, and uses the Gibbs sampling method to update the current parameters with fixed other parameters to assign initial values to the variables. Then stochastic optimization method is used to optimize expectation maximization (EM) optimization for two kinds of variables. The posterior distribution of the dictionary atom is obtained by minimizing the Kullback-Leibler (KL) distance, and the dictionary size is derived non-parametrically. Finally, the image to be reconstructed is divided into smooth and texture patches by bilateral filter during reconstruction, the sparse reconstruction method is used for the texture part while the bicubic interpolation reconstruction is applied for the smooth part. Compared with the bilinear, the bicubic interpolation and the super-resolution reconstruction algorithm based on sparse representation, the average peak signal-to-noise ratios of the proposed method are increased by 3.85, 2.11, 0.20 dB, respectively. And the average relative global dimensional synthesis errors (ERGASs) are decreased by 0.64, 0.28, 0.04 dB, respectively. Experimental results show that this algorithm can provide more high-frequency detail information by adding more sample and parameter prior information, which has certain universality and strong noise robustness, and the reconstruction speed is faster.

Enhanced self-heterodyne synthetic aperture imaging ladar (SAIL) with a transmitter of two coaxial polarization-orthogonal beams, uses heterodyne receiver to receive echo signal, self-heterodynes the two polarization-orthogonal echo signals during signal processing to eliminate the phase error generated by atmospheric turbulence, mechanical trembling and ladar system. A laboratory demonstration of enhanced self-heterodyne SAIL for two-dimensional (2D) target is proposed, the data collection and the image processing of this demonstration is mathematically detailed, and its imaging experiments for 2D target are also presented. The achieved imaging resolutions are 3.6 mm×3.6 mm, and two point-array targets are also reconstructed with high contrast. The experimental results prove that the enhanced self-heterodyne SAIL is feasible for 2D imaging.

Numerous registration methods for airborne and terrestrial light detection and ranging (LiDAR) point cloud utilize geometry information of three-dimensional point cloud. Corresponding features of airborne and terrestrial LiDAR point cloud are matched, and point cloud coordinate transformation parameters are calculated to realize point cloud registration. A new registration method based on laser intensity classification is proposed. Firstly, the laser intensity of airborne and terrestrial LiDAR point cloud is corrected and classified. Then, the plane features are extracted by the classification results. The corresponding plane features are matched taking topological relationship and the classification results as constraint conditions. Finally, the coordinate transformation parameters are calculated to register the airborne and terrestrial LiDAR point cloud. The results show that compared with traditional methods, the proposed method can reduce registration errors from differences of the scanning angle and density between airborne and terrestrial LiDAR. The proposed method can still achieve accurate registration effect when the geometry shapes of the corresponding features of airborne and terrestrial LiDAR are not completely identical.

Airborne LiDAR point cloud filtering is a key step in point cloud processing. Lots of experts and scholars at home and abroad are doing research on point cloud filtering. In recent years, filtering is developed very fast and many other algorithms based on new theoretical background are proposed. Thus, it is urgent to summarize all kinds of filtering algorithms systematically. We classified all the algorithms into six categories based on the previous studies. The principles, implementation steps and existed problems of each class were also elaborated. This paper adopted the data sets provided by the International Society for Photogrammetry and Remote Sensing (ISPRS) to compare the accuracy of each representative algorithm in each class and summarized their advantages and disadvantages. Last but not least, we provided some advices on how to further improve the accuracy and robustness of filtering algorithms. The review will be beneficial to point cloud data processing researchers to have more systematic, clear and accurate knowledge on filtering algorithms. It is also expected that this paper would make some contributions on extending filtering algorithms and improving point cloud post processing precision.

Fano resonance, theoretically explained by U. Fano, is a type of resonant scattering phenomenon in the quantum system that results from interference between discrete and continuous state energy, and gives rise to an asymmetrical line-shape. In recent years, plasmonic Fano resonance has been found in the plasmonic structure system, which is produced by the interaction between the superradiation mode and the subradiation mode supported by the structure. Because the Fano resonance has narrow frequency linewidth, weak radiative losses, and strong near-field enhancement on the surface of structures, it has attracted much attention in the field of photonics. The disk-like structure which has the ability of supporting wider superradiation linewidth can couple with one or more subradiation patterns supported by the structure to excite and modulate single or multiple Fano resonances effectively. Moreover, the disk-like nanostructured systems can still excite high-intensity plasmonic Fano resonance under the case of highly geometric symmetry or regularly multi-individual composition, which can further expand the design of Fano resonant nanostructures. In this paper, we summarized the disk-like Fano resonance nanostructure, including single disk structure, heterogeneous dimer disk structure and multipolymer disk structure, and presented the mechanism and excitation of the plasmonic Fano resonance. In addition, we also briefly discussed the applications of disk-like Fano resonance nanostructure.

In the systems of micro positioning, navigation and timing (Micro-PNT), the chip-scale atomic clock (CSAC) serves as the core of physical micro-clock module, whose development has a strong connection with positioning accuracy and timing ability of micro-PNT system. As the key part of CSAC, the atom vapor cell directly determines the volume, stability and power consumption of the atomic clock. With the reduction of power consumption and volume of atomic clock, it is imperative to manufacture corresponding micro atom vapor cell. Taking realization of atom vapor cell processing technology as our targets, traditional glass blowing and micro-electromechanical system (MEMS) techniques are studied in this dissertation. Based on the research of atom vapor cell, this overview summarizes great progress in the past few years and analyzes the shortage of current preparation technology, giving an importance guidance to the preparation technology of vapor cell in the future.

It is urgent at present to realize an all-weather space quantum communication and enhance its communication speed. The development history and the trend of quantum secure communication are summarized, and the development status and the research significance of the free space quantum key distribution (QKD) are also concluded. The source of bit errors and some existing problems of free space QKD are analyzed and the importance of background light suppression is discussed. The suppression methods are introduced and compared,and the next development direction of background light suppression is briefly discussed.

Terahertz time-domain spectroscopy (THz-TDS) and density functional theory (DFT) simulation are adopted to study the spectral properties of arabinose chiral isomers in the terahertz band. It is found that L-, D- and DL-arabinose show their characteristic absorption peaks in the effective spectrum range of 0.4-1.8 THz, and the absorption spectra of the three arabinoses are obviously different. The absorption spectra of L-, D- and DL-arabinose are calculated with the use of density functional theory in Gaussian09 software, and the characteristic absorption peaks are identified. The results show that the theoretical simulation is consistent with the experimental result and terahertz time-domain spectra can be used to identify the arabinose chiral isomers and their racemic compounds. This indicates that the correlation between chiral isomers and their terahertz spectra can be revealed by terahertz time-domain spectra. So terahertz time-domain spectra provide a new method for studying the detection and analysis of chiral isomers and their racemic compounds.

Compared with the traditional solid state lasers, the fiber lasers is conducive to the miniaturization of devices and the promotion of laser induced breakdown spectroscopy (LIBS) technology. In this paper, the fiber lasers LIBS (Fiber-LIBS) technology is applied to grade identification of aluminum alloy. The data classification, normalization, support vector machine, and principal component analysis are used to classify the grades of 24 samples of 6 kinds of aluminum alloys. The results show that, compared with the simple classification algorithm based on the support vector machine classification algorithm, the data filtering, normalization, and support vector machine combined with the principal component analysis can make the average prediction accuracy rate increase from 92.34% to 99.83%, and can decrease the modeling time more than one order of magnitude. The experimental results show the feasibility of fiber lasers used in LIBS system for the metal grade recognition.

To achieve the fast discrimination of five varieties of honeys, namely linden honey, vitex honey, rape honey, acacia honey and litchi honey, we propose a new method in this article by using the mid-infrared spectra based on principle component analysis (PCA) combined with linear support vector machine (SVM) or least squares support vector machine (LSSVM). The mid-infrared spectra of five varieties of honey samples are determined by Fourier transform infrared spectroscopy and normalized. Then the 5-dimensional, 10-dimensional, 15-dimensional, and 20-dimensional feature data will be extracted from spectra with the use of dimension reduction method of PCA after normalization. Finally, the two classifier models, linear SVM and LSSVM with radial basis function (RBF) based on the grid search optimization, are designed. Using different classifier model, we identify the different dimensional feature data extracted from spectra data of unknown honey samples. Then the results of different dimension feature data and different support vector machines are validated. Experimental results show that for the 20-dimensional feature data obtained by the dimension reduction method of PCA, an average recognition rate of higher than 97% on SVM and LSSVM classifiers is achieved, the highest recognition rate can reach 100%, and classifier stability is very good. LSSVM classifier has higher recognition accuracy and better stability than linear SVM classifier in classification with lower dimension data. Hence, it proves the feasibility of rapid identification of five varieties of honeys with mid-infrared spectra combined with linear SVM or LSSVM.

The temporal performance of framing tube can be effectively promoted by pulse-dilation technology. However, the pulse jitter causes the exposure time jitter, which reduces the reliability of the tube. In order to discuss the reason and the improvement method of the exposure time jitter, the measurement, errors analysis and modification of exposure time are carried out. The results of the analysis of 200 dynamic images show that the jitter scope of the exposure time, caused by pulse jitter and inconsistency of dilation slope, is in the range of 11-26 ps. And the evaluation value of the exposure time of the tube is about 17.3 ps by using weighted means method according to the characteristics of dilation pulse and the distribution ratio of exposure time. Using the initial amplitude of the stretching pulse to correct the synchronization voltage, the deviation between the measurement value of the exposure time measurement and the theoretical value is reduced from ~9.8% to ~1.7%, which effectively increases the credibility of the measurement value. The study results provide the theoretical reference and technical support for the improvement of the reliability of the exposure time of the tube.