In consideration of the effect of stray light on the measurement accuracy of a space-borne differential optical absorption spectrometer for monitoring atmospheric trace gases, several stray light removing mechanisms including a light shield and other baffles were developed to suppress the stray light, and the properties of the stray light were analyzed. With TracePro model, the stray light removing of UV channel 1(240-315 nm) in the system was evaluated ,then the first and second order scattering paths were determined by simulation analysis. On the basis of stray light paths, the stray light removing for revised system was analyzed, and Point Source Transmittance(PST) was calculated. The obtained PST value shows that the light shield and other mechanisms have suppressed the stray light effectively, PST values are less than 3×10-5, and stray radiation ratio in a central field is 5.472×10-4. Furthermore, the final stray light suppressing level of the system satisfies the requirement of design index. Moreover, the edge filter method is used to measure the stray light level, and the results indicate that the stray light in the central field is 8.167×10-4, which is close to the simulation results and proves the simulation result to be credible.

In order to calibrate the six apertures of a multi-angle polarizing radiometer synchronously and to meet the need of full beam path calibration. a six-hole ellipsoid integrating sphere calibration source was designed according to the optical characteristics and structural constraint of the polarizing radiometer. Then the consistency, uniformity and polarization of the integrating sphere output were tested. In the test, the consistency was directly measured by using the scaled spectral radiometer, while the uniformity was detected by using the scanning method combined a lens and apertures in consideration of the small outlet of the integrating sphere. Furthermore, the degree of polarization was detected by a linear polarization detector designed based on the Marius extinction law. Experiments demonstrate that the detection methods and data processing have enhanced the signal-to-noise ratio as well as the accuracy of the system. It is shown that the brightness consistency of paired openings for the integrating sphere is better than 99.5%, the surface uniformity is better than 99%, and the degree of polarization is below 1%, which satisfies the needs of radiometric calibration of the multi-angle polarizing radiometer.

To realize the calibration of a camera and a light plane in the line structured light sensor simultaneously, a calibration method was proposed based on a single plane calibration target which contains a concentric circle and two orthogonal lines passing through its center. Firstly, several images of model plane were captured from different orientations, and the camera parameters were determined via computing conjugated points linearly and were optimized according to the orthogonal constraints. Then, by moving the sensor for several times freely and keeping the light plane intersecting with the concentric circle, the perspective-three-points (P3P) model was implemented, and the 3D coordinates of calibration points on the light plane were computed. Finally, the light plane equation was derived eventually using the least square algorithm. Experimental results indicate that this calibration method has high accuracy, its average measuring accuracy is 0.036 82 mm, and relative error is about 0.277 13%. This calibration method requires only one target to complete the calibration of camera parameters and light plane equation, therefore, the calibration cost is reduced. It is characterized by simple calculation, easy operation and suitable for calibration in fields.

The assembly variables of a high performance optical system are dependent on the design of optical and mechanical structure, however, the reference coordinates used in the system assembly are mostly different from those used in optical design. To describe paraxial image motions due to adjustments precisely, the Gaussian optical homogeneous coordinate transformation model with assembly error variables was established under a reference coordinate. According to specified optical-mechanical design, the Gaussian image rotation and defocus as functions of assembly variables were described. Then, the paraxial image motion induced by small deformation of a three-mirror off-axis telescope was calculated, which shows a relative difference less than 4％ compared with that from the optical software optimized image location. By the variance combining method, a linear optimization model was solved to get the loosest error budget for 17 variables and the Monte-Carlo simulation was used to verify the error budget. It indicates that all fields meet the focus depth of ±10 μm within the focusing ability of ±300 μm. This method ignores subtle influences caused by aberration, and is favorable for optical systems consisting of plane optical components and near aberration-free components.

As the aperture for a monolithic mirror was limited in 1.2 m by traditional technologies, a new reaction-formed joint technology for RB-SiC was proposed. With the proprosed technology, the SiC mirror was joined in the green body process, and the green body densification and the joint of mirror were finished in the same sintering process. A 230 mm diameter RB-SiC mirror was fabricated by this technology. After milling and polishing finely with the FSJG-2 facility, the figure error of the mirror surface is less than λ/50(λ=632.8 nm) . The mirror was tested in the temperature range of (20±3) ℃, and the tested results show that the change of the figure error of the mirror surface is less than λ/300, and the mirror surface figure is not changed obviously after the thermal cycle test. Moreover, the roughness of the surface near the joint line is 3.3 nm, and its microstructure is similar to that of the RB-SiC ceramic. The thermal property of welding line is matching with that of RB-SiC ceramic. Obtained results demonstrate that the reaction-formed joint technology for the RB-SiC mirror satisfies the need of the large aperture mirror used in space optics.

An scanning surface plasmon resonance biosensor was constructed to quantitatively detect array samples. First, based on the optimal rotation position in plane prism-coupling mode and a double-prism detection optical path, the scanning surface plasmon resonance biosensor for array samples was constructed. Then, the sample density of array samples was calculated. Finally, by taking the distilled water and glucose solution with different concentrations in 5, 10, 15, 20, 25, 30 and 35 mg/mL as array samples to be measured, the multi-sample spots were detected by surface plasmon resonance method. Experimental results show that the resonance angles of array samples are 73.745, 73.919, 74.052, 74.185, 74.306, 74.408, 74.549 and 74.660°, respectively, and the solution concentrations and their resonance angles show a good linear relation, which proves that the proposed method and device are feasibility. The device is significant for increasing the sample density of array samples and realizing the high precision detection.

A novel heterodyne interferometric ellipsometer was investigated to measure the thicknesses of thin films in the nanometer-accuracy. The depolarization effect of a Nonpolarizing Beam Splitter (NPBS) was characterized by the transmission-induced-retardance (TIR) between p and s components of a polarizing beam, reflectance ratio, transmittance ratio and reflection-induced-retardance (RIR),then a corresponding error model was established. The influence of the depolarization effect and misorientation of a multi-layer dielectric NPBS on the errors of ellipsometric parameters was investigated. Experimental results indicate that the fluctuation of the depolarization effect caused by the environmental temperatures, incident angles and the change of polarizating state has a major impact on the measurement accuracy, and it can not be eliminated by calibration. In order to implement the nanometer-accuracy, the misorientation angle must be smaller than 0.1°. Furthermore, it shows that the NPBS orientation angle errors for recombination of the measurement light and the reference light have more important contribution than another ones to the measurement accuracy. The thickness measurement error of thin film induced by NPBS is about 1.8-2.5 nm，which means that the NPBS is an assignable error source in the Mach-Zehnder interferometric ellipsometer.

To analyze the principle of profile formation for grating masks and the evolution of photoresist response curves, a simulation model of profile formation for grating masks in development was established. Based on the difference of photoresist dissolution rate in the different regions, the complete photoresist curve was divided into three sections, the effect of each section in the profile formation of grating masks was analyzed, then the simulation surface-relief profile model was presented. The experimental results indicate that the groove profile inclines to be rectangular or trapezoidal when the nonlinearity of photoresist response curve is remarkable, and the groove depth is mainly decided by the initial photeresist thickness. The groove profile is sinusoidal when the linearity response is strong, and the groove depth is also always decreased under this condition. The experiment shows that the proposed model can predict the profile evolution for the different photoresist curves and it provides a directive theory for fabricating the various profile masks during development according to the different photoresist response curves.

Based on a plastic scintillator and an optic streak camera, a measuring system for fusion reaction-rate was designed, then the working princile and design method of the measuring system were addressed in detail. A piece of fused silica window was set just behind the plastic scintillator to protect optical elements from darkening after a long time exposure in scatted X rays and the ultraviolet crystals and other ultraviolet glass were selected to ensure the high system transmission. Furthermore, the image plane of the optical system was put onto a photocathode of the streak camera directly to remove the aperture non-matching of input optic system of the streak camera. Finally, a field mirror was put on the first image plane to reduce the max diameter of optical element from 160 mm to 120 mm. The designed optical system has a total length of 2 660 mm, magnification of 3∶1 and its final lens is F/0.667 and transmission is 67%. Moreover, the time dispersion of the optical system is less than 7.3 ps. These results meet the experimental needs of different neutron yields and good experimental results have been obtained in a laser targeting experiment.

Developments of Deep Ultraviolet(DUV) coating technologies in Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences(CIOMP) were reviewed. Some researches were carried out systematically to meet the requirements of high quality DUV optical systems for coating processes. Two special systems were customized for evaporation and ion beam sputter deposition techniques respectively, which could coat the substrates for φ410 mm optical elements. After optimizing deposition technology, the typical transmission for a both-side coating sample at 193 nm is in the range of 98.5%~99%. The factors that effect on the figure accuracy of optical elements were investigated, the obtained thickness uniformity of coatings is 0.1% rms and it meets the required tolerances of high quality DUV optical system. Coating stresses were measured by using X-ray diffraction technique and their effects on the figure error were analyzed by using finite element simulation. In consideration of the practical utilities of coatings, several kind of solving methods were proposed such as the recovery of the transmission degradation caused by environmental pollution with UV radiation and the accurate control of coating thicknesses based on crystal monitor technique. Finally, it points out the research directions in further based on the results mentioned above.

A alternative magnetic field generator is prepared by special circuits and active coils and the effects of different magnetic field intensities on the macro morphologies and microstructures of Fe-based composite coatings are studied by laser cladding process. Based on the electromagnetic principle and liquid metal solidification theory, it describes the solidification process of laser cladding coatings and reveals the main mechanism that magnetic field induced columnar dendrites transform into equiaxed crystals. The results show that the skin effect of liquid metal molten pool surface and alternating electromagnetic force allow the solidified cladding layer surface morphology to the wave form. The height of cladding layer and cross-section area decrease with the increase of magnetic field current, but the cladding layer width changes a little. The electromagnetic force in molten pool interior drives the dendrite ablation and mechanical break, and the free broken dendrites become a new nuclei to increase the nucleation rate, by which the top microstructure of cladding layer transforms from dendrite crystals to equiaxed crystals. Furthermore, the equiaxed crystal zone becomes wider with the increase of magnetic field current, but the bottom microstructure has a little change.

An oxygen optical sensing probe was fabricated by modifying the sensitive film doped with Ru compounds in a holy optical fiber. The structure of this probe was based on the simple holy optical fiber with internal microchannel structure throughout the whole fiber, in which the microchannel structure could be used as a carrier of sensitive materials and a micro-sensing pool and the sensing process was finished in it. The sensing film was deposited in the optical fiber by inhaling the cellulose solution containing ［Ru(dpp)3］2+ directly. Furthermore, the sensing process was based on fluorescence quenching process. As the oxygen could strongly quench the fluorescence of ［Ru(dpp)3］2+ in the sensitive film, the sensing probe could exhibit different fluorescence intensities for different oxygen concentrations. The experimental results show that the designed probe responses to the oxygen in linearity for a full concentration range of 0～100% , and the ratio of I0/I100 (where I0 and I100 respectively represent the fluorescence intensities of the optrode exposed to 100% N2 and 100% O2) is 9.4. Simultaneously, the probe can make a quick response within 50 ms. The experiment also show that the error between experimental value and real value is less than 3% and the resolution can be 2% in a normal pressure.

The high accurate control system for a diffraction grating ruling machine was designed by taking the S3C2440A as a control core to implement the real-time shows, real-time control and other functions. The S3C2440A micro controller was used to control a servo motor and a stepping motor to complete the indexing movement and the ruling movement of the grating ruling, the SiGNUMTM RELM straight line grating' feedback and PD plus feedforward control algorithm were taken to compensate the index error caused by the inertia and other reasons, then the indexing movement using PD plus feedforward control algorithm was simulated by MATLAB. Furthermore, the error caused by the out-of-step of stepping motor was compensated by the feedback of a RON225 incremental angle encoder, and the best speed of the indexing movement and the overshoot distance of the servo motor when it stopped were collected by the analysis and the process of the experimental data. Finally, the analysis on the controlling error was introduced, and how to improve the system in further was put forward.

The gimbal servo system for a Control Moment Gyroscope(CMG) directly driven by a Permanent Magnet Synchronous Motor (PMSM) was designed, and an Active Disturbance Rejection Control( ADRC) algorithm with integral feedback was proposed. Firstly, the mathematic model for the CMG gimbal servo system was built by using an oriented control strategy in shaft current id=0. Then, the effects of both frictional and alveolar torques on the performance of gimbal servo system were analyzed. A simulation system for gimbal servo system which used the ADRC as a speed loop was built up in Matlab. Finally, the hardware experiments of fuzzy PI, ADRC and the ADRC with integral feedback were carried out. Test and experiments show that the steady state accuracy is 0.005-0.012 rad/s when a step velocity of 0.1-2.0 rad/s is tracked by using the ADRC., and that is 0.001-0.005 rad/s and the crawling speed is 0.003 rad/s when the step velocity of 0.0-0.1 rad/s is tracked by using the ADRC with integral feedback. Furthermore, the relative amplitude error is 0.55% and the phase error is 0.099 78 rad, when the gimbal system tracks the 2sin(t) rad/s by using the ADRC with integral feedback. The proposed gimbal servo system of CMG satisfies the demands of high precision and robustness.

A displacement detection method based on two-wavelength grating interferometry was presented. A measurement experiment was carried out by using an integrated grating sensor chip fabricated by silicon-glass bonding bulk process. The measurement setup consisted of a sensor chip, two lasers with wavelengths of 640 nm and 660 nm, two photodetectors and a data acquisition circuit. The sensor chip included moving parts with a reflective surface and a metal grating on the transparent substrate. In experiment, the laser beam illuminated the grating and the grating reflected the laser beam to form a diffraction pattern. As the intensity of the diffracted beams changed according to the displacement between movable structure and substrate, the displacement could be determined with an extended range than single wavelength and the absolute position could be obtained by measuring intensity signals of the two wavelengths respectively. Experimental results show that the initial gap between the moving part and the substrate is about 7.522 μm measured by proposed method and it can implement the displacement measurement in a range of 618 nm with a resolution of 0.2 nm.

The compensation accuracy of a forward image displacement compensation device in an area scan color CCD camera was analyzed, because it would influence the resolution of the camera directly. The characteristics of the mechanic compensation device were researched, and the allowable residual error of the velocity compensation corresponding exposure time was given. A mathematical model for the residual error of velocity compensation was established after analyzing main error sources of the device, then the error model was simulated by the Monte Carlo method. Finally, the compensation accuracy was verified by an imaging test with a dynamic target simulator. The test results show that the residual error of velocity compensation is 395.6 μm/s, which is coincident with the simulation data 400 μm/s. It is concluded that the method can estimate accurately the matching error accuracy.

To solve the problem of inaccurate peak demodulation resulted from the system uncertain phase shift in Amplitude Modulation (AM) for a capacitive displacement sensor, an AM capacitive displacement sensor based on an improved delayed feedback peak detector circuit was proposed and its applied functional blocks and circuits were investigated. Firstly, the measurement principles of the AM capacitive displacement sensor and its detection circuit were analyzed. Based on generation principle of phase shift with the AM signal, an improved delayed feedback peak detector was presented against the problem of inaccurate peak demodulation. Then, the AM capacitive displacement sensor was prototyped using printed circuit boards and its performance indicators were tested as well. Finally, experimental results and errors were discussed. It indicates that the linearity of the delayed feedback peak detector circuit is better than 0.05%, the data shift of prototyped sensor is less than 10 nm/30 min, and the maximum deviation of sensor is 36 nm within the measuring range of 0~25 μm. It concludes that the delayed feedback peak detector circuit can effectively solve the problem of inaccurate peak detector resulted from the system uncertain phase shift, and the proposed AM capacitive sensor can satisfy the requirements of high precision measurement.

A novel Hex-Rotor aircraft was proposed to overcome the effect of under-actuation and strong coupling characteristics on the flight performance of existing multi-rotor aircrafts. Based on the unique configuration of six driving rotors, the Hex-Rotor aircraft has the ability to control the space 6-DOF channels independently . First, the structures and characteristics of the Hex-Rotor aircraft were introduced and its dynamic model was established. Then, an autonomous flight control system with a parallel double-loop structure was designed, in which the command-filter backstepping approach was introduced into the attitude stability augmentation control loop, and the translational controller based on the active disturbance rejection control technique was used in the position loop. Finally, the total independent control of the aircraft on 6-DOF channels was achieved in the numerical simulation and the influences of unknown external disturbances and model uncertainties on the flight control performance were overcome. The prototype experiment results indicate that the horizontal tracking errors, altitude errors and attitude errors for the aircraft are limited in ±4 m, ±3 m, and ±005 rad, respectively, which are all according with the precision ranges of measurement units and mean that the prototype has tracked the reference translational and attitude commands accurately. The simulation and experimental results verify that the designed Hex-Rotor aircraft has desired maneuvering capability, and the control system is able to guarantee the autonomous tracking flight of the aircraft.

As the surface damage detection plays an important role in evaluating engineering ceramic surface quality, this paper introduces an image reconstruction algorithm, Nonnegative Matrix Factorization (NMF) algorithm, into the damage detection of engineering ceramics grounding surface for the first time. It analyzes the theoretical function of the algorithm and gives a detection example. First, the input image data set was reduced from an original data space to a lower-dimensional NMF space, and the image reconstruction relative error 0.1 rule proposed by this paper was used to determine a proper space basis r value. Then, the background image of ground texture was obtained by image reconstruction using two lower-dimensional nonnegative matrixes, and the ground textures were removed by image subtraction. Finally, the Canny edge detection was used to extract the damage image of engineering ceramics grounding surface. Experimental results indicate that the proposed method can accurately extract the surface damage of engineering ceramics and can calculate the evaluation parameter of grinding damage rate.

A three-dimensional thermal-mechanical Finite Element Model (FEM) was established by the ANSYS code to reveal the influence of fixture restraint distance on Pulsed Laser Welding (PLW) distortion of Hastelloy C-276 thin sheet with a thickness of 05 mm. The influence laws of the restraint distance on welding longitudinal deflection and transverse shrinkage distortion were investigated by using the displacement restriction assumption instead of the actual restraint condition. Experiments were carried out to measure the welding thermal cycle curves and residual distortions. The simulation results are in good agreement with that of the corresponding experiment, which validates the reliability of the FEM. It shows that the welding longitudinal deflection and transverse shrinkage distortion increase linearly with the restraint distance increasing from 8 mm to 20 mm, especially the welding transverse shrinkage distortion increases by nearly 3.6 times. Furthermore, the welding residual distortions can be controlled by reducing the restraint distance. Results dennonstrate that the FEM can simulate accurately the temperature field and the residual distortion of PLW and can provide theoretical and experimental basis for suppression of the welding distortion of thin sheets.

According to an actual engineering project and dynamic topographic features of moving workpieces, a new measuring model for large scale workpieces was established based on the virtual measurement datum structured by the two-dimensional laser triangle measuring method and the multi-sensor fusion rule. Then, a noncontact, high precision and low cost dynamic length measuring system based on the human-computer interaction was developed by using the error separation method to correct theoretical errors and motion errors. In the experiment, the two probe units were placed on the two sides of moving workpiece, the displacement between both end planes of workpiece and virtual measurement datum was measured simultaneously. Finally, A experimental measurement on the cylinder workpieces (1 000±25) mm at different speeds was performed, and results show that the system can offer the resolution less than 10 μm, measuring precision over 100 μm, and a maximum speed of 5 cm/s. According to the actual operation results, the proposed measurement techniques and measuring system can be well applied to the length detection for similar regular large scale workpieces in industrial production.

To correct the measuring angle errors derived from the eccentricity of installing a circular grating in a manipulator joint test system, a model for compensation of grating measuring angle errors and the test method for installing eccentric errors are investigated. First, according to the relationship between angular error and eccentric parameters, the paper deduces the model of angular errors caused by install eccentricity. Then, it describes the method and principle to detect the grating eccentricity and eccentric direction by receiving phase errors of signals using two sensors in contrast. Based on the Lissajous plot generated by composite signals, the eccentric parameters of the grating in manipulator joint test system is measured. Finally, using the angle error formulas to compensate the measuring angle, the measuring system has been amended. The experiment shows that the amended angular accuracy has greatly improved, and the measurement accuracy is increased by nearly five times. It can satisfy the requirements of high accurate systems.

Orthogonal tests in two rotational modes (the rotational tool mode and the rotational worktable mode) are conducted to confirm the synergistic effect mechanism of an electro-magnetically coupled field in the Electro-Magneto-Rheological(EMR) effect-based tiny-grinding wheel finishing. Under the electro-magnetically coupled field, the EMR polarized particles of the rotational tiny-grinding wheel generate the Lorentz force and the spin couple, then the in situ vibration of the EMR particles due to the spin couple can impact regularly on the workpiece surface and promote material removal, so the material removal depth in the rotational tool mode is larger than that in the rotational worktable mode. However, if the interaction force of the particle chains is weak, the in situ vibration will break the particle chains and reduce the material removal. The couple mode of electric and magnetic field has a significant influence on the machining efficiency of the EMR finishing. It shows a good synergistic effect of the EMR finishing in the voltage of electric field of 1 kV and the excitation voltage of magnetic field of 5 V in the rotational tool mode.

The image motion of an aerial imaging system was researched and its effects on imaging quality and the resolution of the camera were analyzed. In order to obtain and compensate the image motion of the aerial imaging system, a computation model of image motion based on coordinate trans-formation was proposed. The coordinate transformation model from a ground object to a sensor plane was established by using linear coordinate transformation, and the computation formula of image motion was derived based on the different coordinates of the same ground object during camera exposure. The sources of image shift error of the imaging system were analyzed and the influences of various error factors including the aircraft motion, attitude angle and the camera swing angle on image shift calculation were discussed. Then, the calculation error budget and synthesis with Mont Carlo method was presented. The experimental result shows that the image motion is directly proportional to the velocity of aircraft, and inversely proportional to the distance between aircraft and ground object when the attitude angle and camera swing angle are constant. Furthermore, the calculation error becomes greater with the increase of the parameter disturbance, especially the measuring errors of aircraft motion in longitude and latitude. It concludes that this method is efficient and will be useful to the image motion compensation of aerial imaging systems.

A method combined the modified Lucas-Kanade Gait Flow Image (LK-GFI) with the view was proposed to solve the problem that personal identification based on a gait is sensitive to view change. The Lacus-Kanade optical method was used to compute the optical flow between two silhouettes to construct LK-GFI, and the view was obtained according to the walking direction of the person. The LK-GFI database for the target at different views was established, then the new person's view and LK-GFI were extracted. At last, the similarity between the new persons LK-GFI and the targets LK-GFI at the same view was computed by the Euclidian distance method. The performance of this method was evaluated on the data in the CASIA database and the data obtained in indoor lab environment, and the False Rejection Rate (FRR) is 7.95% and 9.12% respectively. It is reduced by 12.5% and 14.45% respectively compared with that of the Gait Energy Image (GEI), and by 7.77% and 6.74% respectively compared with that of the Gait Flow Image (GFI). The proposed method has high recognition accuracy, strong real-time and the robustness to view changes.

To improve the precision, update rate and the anti-noise ability of star sensors, this study was focused on the extraction of star images. Firstly, the criterion to confirm whether a pixel attributes to the same star coverage with the local maximum pixel was established according to the gray distributing features of star images. Then, several kinds of methods for image extraction were introduced, such as the block division of imaging array, and the prediction of background noise. Finally, by regarding the peak value pixel as the origin of the circular region growing, the local region growing criterion was set up successfully based on the gray distributing features of star images. The simulation conducted in the case of no noises shows that all of the simulated star images in the reference star map can be extracted successfully, and the sub-pixel location accuracy is 0.028 2 pixel by using centroid method. Moreover, under upmost adverse condition with a Gaussian noise mean value of 20 and standard variance as high as 2.5, the success extraction rate can still reach 86.11% with a decreased centroiding accuracy of 0.219 6 pixel. The test about a faint star image in a star map with poor uniformity and a low SNR of 4.9 dB also proves the excellent detection ability of the proposed method, and it shows advantages of good real-time property and high correct extracting rate for star images.

A Method of Counting Distances (MCD) was proposed to improve the efficiency of iris recognition in this paper. The MCD was used in a denoising for iris edge image, and efficient location methods for both inner and outer circles of the iris were established. First, a method of grey level summation was utilized to find a pixel in the pupil, meanwhile, the iris edge image could be obtained by using the Canny operator, and three edge points of iris inner edge could be found according to the pixel in the pupil. Then, three edge points were derived to the equation of a circle to find the iris inner edge by substituting. Furthermore, the noises in iris outer edge image were eliminated in terms of the inner location data, and 20 percent of the edge points far from the center of inner circle of the iris were excluded from the denoised edge image. Finally, an improved Hough transformation was used to obtain the data of iris outer edge. Experimental results show that both inner and outer location time obtained using MCD is lower than those obtained by the Fast and Accurate Iris Location Algorithm (FAILA), and the success rate of MCD is 96.67%, which is higher than that of FAILA. Thus, the MCD has high precision and fast speed, and can lay the foundation for improving the efficiency of iris recognition.

A single pixel camera was used to shoot different scenes by taking a sparse binary matrix as the measurement matrix. The single pixel camera based on compressed sensing theory could acquire a high resolution image by using a Digital Micromirror Device (DMD) and a single detector. It captured and compressed the picture simultaneously, so that the data amount, scale, complexity and the cost of the system were reduced greatly. In this paper, the effect of measurement matrix upon the precise of reconstructed picture was analyzed, and the condition to reconstruct the binary matrix for object picture precisely was given. An experimental platform was established and imaging experiments under three different scenes were performed for different strokes of the words “China” and complex real objects using sparse binary measurement matrix. The picture with the picture resolution of as large as the size of the DMD was reconstructed by gradient projection algorithm. Experiment results indicate that single pixel camera could reconstruct the image precisely when the number of measurements are 20%~30% of the total number of pixels for the object image. It concludes that the single pixel camera can acquire high resolution images.

To match automatically rotated stellar images, a rotation invariant matching method based on invariant feature descriptors was proposed, in which the Speeded Up Robust Features (SURF) was used to describe and match star features for the first time. First, a stellar image was segmented, and the non-maxima value was suppressed to extract star points in the stellar image. Then, a star distribution scale factor was calculated, the dominant orientation was obtained in a circle region with a radius of 6s, and the 20s×20s local region was rotated to the dominant orientation. In the local region, the SURF descriptor was calculated for each star. Finally, an automatic matching strategy based on the difference between dominant orientations was proposed. By this method, the threshold was calculated automatically and the transform matrix was given. Experimental results demonstrate that the proposed method can robustly detect star features and achieve a high precision stellar image matching between images with rotation, translation and perspective change. Obtained results show that correspondent star errors is below 1 pixel and 15 pixel for simulation and real image experiments, respectively. It indicates that the method to apply SURF descriptor to star matching and recognition is feasible.

To achieve the robust tracking for a visual object under challenging conditions in the noisy, occlusion and the deformation, a novel visual object tracking method is proposed in this paper. By combining the Kernel Sparse Representation Classification (KSRC) and adaptive dictionary updating method under Normal Hedge framework, this method can handle tough situations like high inter-class similarities and drastically target appearance variations. Although the KSRC enhances classification performance, standard convex optimization is not fast enough for tracking in real time. Thus an efficient Kernel Random Coordinate Descent(KRCD) method is proposed to calculate the sparse coefficient vector, and the KRCD-SRC classification method is taken to calculate the loss value of each particle. In order to avoid the template drifting, the adaptive dictionary updating method is also given. At last, the states of the target are estimated by the Normal Hedge. Experiments show that the average computing frame rate of the proposed method is 14 frame/s when 50 particles are used. Extensive test results suggest that the proposed method outperforms several state-of-art tracking methods in many complex conditions.

An error correction algorithm for CCD images was proposed to improve the reliability of image optical fiber transmission in a Time Delay Integrated(TDI)CCD camera with a large field of view. First, the image optical fiber transmission characteristics for the space camera and data transmission errors in a fiber channel Additive White Gaussian Noise(AWGN) model were analyzed, and a (16, 8) error correction algorithm was proposed. Then, the (16, 8) algorithm idea and error correction principle were described and an implementation method using Very Large Scale Integration(VLSI) and an encoder were explained. Meanwhile, the feasibility of the algorithm for space camera was explained in different aspects. Finally, the experiments were performed for the transmission system of a prototype in a space multispectral camera. The results show that the (16, 8) algorithm has strong error detecting ability, less resource occupied rate and easier to realize with a hardware. The method is able to correct 191 bit errors in 3 043 Byte, and the resource occupied rate for the encoder is less than 5%. In conclusion, the algorithm improves the reliability of image optical fiber transmission in space cameras.

For a variety of simulation needs of the CCD camera combination in a remote sensing satellite and for three different simulation tasks in ground electronic measurement systems and telemetry data link on board, a type of camera simulation source for different types of CCD cameras is developed with programmable logic devices, high speed array storage technology, Phase Locked Loop(PLL) frequency adjustment technology and the software and hardware programmable image processing technology. The system supports cascading multiple devices, 4-channel parallel 16 bit LVDS, 4 serial LVDS and 4 serial NECL signals, in which the pixel frequency of parallel LVDS data in per channel can be ranged from 0.3 MHz to 200 MHz in a step of 0.1 MHz and the capacity of single cycle output is up to 8 TB. In addition, through a simple configuration, the system can produce the play back data of the remote sensing camera according to the CCSDS protocol and can adjust the output signal frequency and data clock phase in real time. The system has been successfully applied in many types of space camera development tests, and it enhances the efficiency of space CCD camera development, equipment utilization and testing certification means.