To position the target with a large field of view for a novel compound eye system, the design features of the system, mathematical models for positioning target, three-dimensional coordinates of target array, calibration methods and the positioning performance were investigated. Firstly, the design of the compound eye system and its structure were introduced and a mathematical model for positioning the target was established. With an optical splitter,the verticality between target plane and its horizontal movement axle and the parallelity between target plane and compound eye plane were adjusted. Then, By using the methods of vanishing points and the common imaging point of different points on the Y-axis to get the initial point and initial distance, the three-dimensional coordinate of every point located on the target plane was obtained. Finally, by deriving the angles between every target and corresponding channel to achieve the imaging center, the corresponding relation between the angle of each channel and the imaging center was established. Experimental results show that the system calibrated initially can position the target with the viewing angle within 110° on the horizontal plane and 90° on the vertical plane and the distance precision for target positioning can be limited to 2%. The scheme can meet the requirement of compound eye for nonlinear calibration and has operational flexibility.

A novel fiber sensor composed of a Multi mode fiber-Single mode fiber-Multi mode fiber(MMF-SMF-MMF,MSM) structure and a Fiber Bragg Grating(FBG) was proposed for simultaneous measurement of temperature and Refractive Index(RI) based on the interference between guided mode and cladding modes of a Single Mode Fiber(SMF). Due to the different responses of the MSM structure and FBG to temperature and RI variations, the proposed fiber sensor was used to measure the temperature and the RI simultaneously by utilizing a well-conditioned sensitivity matrix equation. The experimental measurement shows that the temperature sensitivity coefficients of the MSM structure and FBG are 0.055 2 nm/℃ and 0.015 8 nm/℃, respectively. The RI sensitivity coefficient of the MSM structure is 109.702 nm/RIU, whereas the FBG is insensitive to RI change. Furthermore, obtained sensing resolution is ±0.32 ℃ for the temperature and ±0.002 3 for the RI. Experiments also indicate that the temperature sensitivity coefficient of the MSM structure is five times over that of the SMS structure. Considering that the cladding mode nature of the SMF is more vulnerable to other parameters of the surrounding medium, the MSM structure can also be applied in other sensing fields.

According to the physical process of flyer plates driven by magnetic fields, an ultra-high speed photographic system including an ultra-high speed rotating mirror streak camera, a micrographic lens suitable for the optical system of the camera and a continuous laser source was developed. An electric-optic switch in series was used to avoid the repetitious exposure on imaging plate of the camera, by which the intense light produced in the course of collision was bated and 1D clear-cut streak images of flyer plates driven by an electric gun were obtained firstly at home. Experimental results show that Ly12 flyer plates with the size of 8 mm×6 mm×0.7/0.9 mm can be accelerated in 9-14 km/s within 1.8 μs when the charge voltages of flyer plate reach in 60-64 kV. The results demonstrate that the charged voltages and the thicknesses of flyer plates have great effect on the movement velocities of the flyer plates. The present work will be helpful for researching flyer velocity and can provide important data for the development of electric guns.

The structure deformation of a micro-spacecraft in orbit caused by launching vibration and its thermal deformation resulted from sunlight will change the optical focal length of an Inertial Stellar Compass ( ISC) in the calibration before launching. Furthermore, it will affect on the high-precision attitude determination for the micro-spacecraft. To improve the calibration, an online fast calibration method based on the attitude matrix criterion was presented. Firstly, this method was used to analyze the mapping relation between the optical focal length and the attitude matrix. Then, combining the attitude matrix generated by the filter in any time with the orthogonal unit features of attitude matrix, the focal length of ISC was calibrated online based on an iterative method. Experimental results indicate that the calibration precision by proposed method for the optical length of the ISC system is equivalent to the extracting precision by the stellar center of mass method, and it is about 0.01 pixel. This method can calibrate the focal length at any time in orbit, and have fast calibration speeds and better calibration results without capturing a large number of other attitude measurement data from sensors.

The actual load-bearing capacities of Power Transmission Towers(PTTs) should be estimated by a real-scale model experiment due to their complex structures, therefore an optical full-field 3D deformation measurement method is proposed to overcome the difficulties of large-scale, large-displacement, 3D deformation, rapid multi-point measurement in PTT real-scale model experiments. Based on close-range photogrammetry technology, this method utilizes specialized markers as measured targets to reconstruct the 3D coordinates of those pre-pasted artificial targets through analyzing the captured photo-set in each epoch, then it conducts the coordinate system global registration according to un-movable targets. By tracking and comparing the 3D coordinates of the deformable targets between different epoches, the 3D load-deformation diagram of the PPT stucture is obtained. Acceptance experiment results indicate that the accuracy of this method could reach about 0.1 mm/4 m and real-scale model experiments show the proposed method could meet accuracy and efficiency requirements of PPT load-deformation measurement. Comparing with traditional deformation sensors, the proposed method can satisfy measuring requirements of non-contact(non-interfere), on-spot, high precision, rapid speed, strong anti-jamming and stabilization.

An adjustable contrast optical target equipment was constructed. After researching the relationship between image contrast and optical contrast, a contrast calibration method by the improved Back Propagation(BP) neural network was proposed. Firstly, the BP neural network model was designed for calibrating the contrast. Then, by combining the Levenberg-Marquardt(LM) with Shrinking-Magnifying Approach, the BP neural network was improved to optimize the convergence speed and generalization ability. Finally, based on the experimental platform of the adjustable-contrast target, the image contrast was obtained by measured radiation data. Comparing with the traditional BP algorithm, the improved one has a better convergence speed and generalization ability. Its calibration accuracy has been improved by 100 times and by 10 times as compared with those of the traditional BP network and the steepest descent method, respectively. When the training times is to be only 2 876 times, the maximum error between calibration value and target calibration value for the contrast is 0.01%, the training mean square error converges is 0.000 459 441, and the test error converges is 0.000 467 003. These results demonstrate that the algorithm is feasible and can meet the demands for contrast calibration in the equipment.

To describe the parameters of a big format Charge Coupled Device (CCD) by a function with definit physical meaning and to evaluate its radiation response performance directly, a concept of “radiation response function matrix” was proposed to describe the radiation response parameters of each pixel of the CCD. Firstly, by analyzing the physical concept of each element, some parameters such as absolute radiate response, nonlinearity, dark noise, Signal-to-Noise Ratio (SNR) and the non-uniformity of each pixel of the CCD were characterized by the matrix. Then, response characteristic coefficient of each pixel was obtained using the model of multiple regression analysis by a test on the area CCD KAI-16000. Finally, by taking the result for an example, application of this matrix was discussed. Experimental results show that the response of this CCD is linear, and its non-uniformity is 3.1%, dark noise is 3.84. These results demonstrate that the method is feasible and practical, and it can satisfy the requirement of the description of a big format CCD.

For the low accuracy of single-level inversion methods to dynamic light scattering, a novel Multi-level Tikhonov regularization inversion (ML-TIK) method combining the Tikhonov regularization method with cascadic multi-grid technique was developed. Firstly, this method divided the original problem into several sub-inversion problems with different grid spaces by a multi-grid technique. Then, from the coarsest scale to the finest scale, each sub-inversion problem was inverted by single-level Tikhonov regularization (TIK) method. Finally, the Particle Size Distribution (PSD) was successively obtained by solving several sub-inversion problems. This method effectively reduces the ill-condition of the original equations. At noise levels 0, 0.005 and 0.01, the simulation data of 200~650 nm bimodal distribution particles were respectively inverted by the TIK and ML-TIK. The results indicate that the inversion PSD of ML-TIK is more consistent with that of the theoretical one and it has better smoothness. Comparing to TIK, the ML-TIK can reduce the peak value error by 8.19% and relative error by 0.448 2. However, when the noise level is 0.005 and 0.01, the PSD of TIK has not obvious bimodal features. Therefore, the ML-TIK has improved the inversion accuracy and noise immunity. Inversion results of 60 and 200 nm experimental data verify above conclusions.

The Signal-to-noise Ratio (SNR) of star-level detection and the energy concentration of a dispersed spot for an lunar-based Optical Telescope(LOT) are tested by using a simulation experiment on ground to verify its detection capability. Unlike the conventional methods that use the effect of CCD hardware parameters on the noise to test the SNR，the new test method directly uses the observation information to compute the SNRs of star targets and the tested uncertainty of object SNRs can be better than 8%. In the test of energy concentration, the centroid algorithm is used to computer the center of the dispersed spot and a Gaussian fitting method is proposed to fit the curve of the energy distribution. The Gaussian fitting method can significantly improve the detection precision of the energy concentration for the dispersed spot, and the testing precision has been improved by 10%. Finally, the detection capability of +15Mv of the LOT is verified by an experiment on the SNR of star level detection and the energy concentration of dispersed spot for the LOT.

An optical system for eyepieces was designed to meet the demand of a wide-angle head mounted display for the wide Field of View ( FOV) , high resolving power and compact , light-weight structure. Twelve groups of high quality eyepieces with a FOV of 33°×24° arraged in a 4×3 array were stitched in a single-eye optical system to realize the design with a wider FOV. The eyepiece was made up of a lens which contained a diffractive surface and an aspheric surface.The dispersion characteristics of binary optical elements were used to correct the chromatic aberration, and the aspheric surface was selected to correct the aberrations( spherical aberration,astigmatism, coma and distortion). Furthermore, an OLED micro-displayer was used to achieve compact and light-weight structure, and the field-of-viewing stitch was used to implement the wide field of view. Experiments show that obtained optical system can offer the horizontal FOV of 120° and the vertical FOV of 60°. Moreover, the resolving power is 43 pixel/degree,the MTF is higher than 0.1 at a spatial frequency of 30 lp/ mm across the entire visual field, and the maximum distortion is less than 3%. The optical system has perfect performance and can meet the domanial demands of virtual reality display and super-wide FOV.

To obtain the smooth surface of an etched-fiber core and to precisely control the diameter of the etched-fiber, an ultrasonic etching system was developed. The effect of ultrasonic power and temperature on the etch rate of fibers (fiber cladding and fiber core) and the surface morphology of etched-fibers were investigated in the 12.5 % percentage concentration of HF solution, respectively. Experimental results indicate that the etch rate is enhanced by the ultrasonic agitation, the relationship between etch rate and etching time is nonlinear and the surface roughness is increased with the etching time in the HF solution. Thereafter, the Buffered HF (BHF) solution was promoted, and the BHF solutions were prepared by making use of 12.5 % HF solution and 25 % NH4OH solution. The influence of BHF solutions on the etch rate and surface morphology was investigated. Obtained results reveal that the smooth surface of etched-fiber core and the linear relationship between etch rate and etching time can be obtained in BHF solution when V (HF): V (NH4OH) is 2, the ultrasonic power is at 165 W and the temperature at 40℃.

As piezoelectric actuators have poor position accuracy caused by their inherent hysteresis nonlinearities, this paper proposed a new modeling method to precisely describe their hysteresis phenomena. Based on the motion rules of hysteresis curves and the nonlocal memory property of the hysteresis nonlinearity, proposed model modified the modeling errors fitted by parabolic model. To verify the feasibility of the model, an experiment was performed by the PST150/7/40VS12 piezoelectric actuator. Experimental results indicate that for the first order reversal signal, the maximum error is 0.141 3 μm and the mean-squared error (MSE) is 0.060 4 μm by using the parabolic model. However, for a more complex signal, those of the parabolic model are 1.396 0 μm and 0.856 6 μm, respectively. When using the amended model to predict the actuator response under the above-mentioned complex signal, the maximum prediction error and the mean-squared error are 0.237 0 μm and 0.09 μm, respectively. These data demonstrate that the proposed model not only provides a minor-loop identical property, but also offers the nonlocal property and it can precisely predict the hysteresis path for assigned complex input profiles.

A Piezo-hydraulic Energy Harvester (PHEH) was presented based on solid-fluid coupling vibration to harvest low-frequency and high-level vibration energy, and its structure as well as working principle were introduced. With an established energy-conversion model, the influence factors on the piezo-hydraulic harvester were analyzed. The theoretic results show that the output performance of the PHEH depends on the vibration frequency/level, the structure and size of piezodisc/cylinder, liquid volume/performance, and system backpressure. The desired performance can not be achieved unless the above parameters are matched well. By a piezodisc with a diameter of 60 mm and a thickness of 0.9 mm and a cylinder with a diameter of 16 cm and a length of 100 cm, a PHVI was fabricated and tested by taking water as liquid medium at different frequencies/backpressures/exciter-amplitudes. The test results suggest that there is an optimal vibration frequency (8 Hz) for the PHEH to achieve the maximal output voltage, which rises with the increasing of backpressure and vibration level. In the case of unvaried other parameters, the achieved voltage from the PHEH under 0.4 MPa is 1.65 times that under 0.2 MPa.

A micro electrostatic seismic power generator based on a droplet was proposed for harvesting the seismic energy in low frequency to improve its low frequency response and impact resistance abilities. The generator structure was based on the charge transfer between the differential capacitors inducted by the movement of the micro droplet in the electric field of an electret. According to the electret model, the charge distribution on the differential capacitors was derived. Combined with the assumption of forced sinusoidal vibration of the micro droplet, a mathematic model of the micro power generator was established. The COMSOL Multiphysics was employed to simulate the micro droplet movement and charge distribution, then the voltage-frequency characteristics were obtained. A prototype of the electret and micro-droplet-based micro seismic power generator was developed and tested. Its resonant-frequency is 3 Hz, the output power is about 0.73 μW on 1 MΩ resistance when the vibration acceleration is 3 g under a resonant frequency. It can scavenge the energies from low-frequency external vibrations.

When the Global Navigation Satellite System (GNSS) receivers are applied to a portable handset, the power of the most power-hungry blocks, such as quadrature 2∶1 frequency divider, should be reduced. Therefore, this paper proposes a 1 V low power quadrature 2∶1 frequency divider at high speed and steady working in all process corners. First, the published high speed architectures are introduced. Then the static DC bias is calculated for the proposed architecture, and a small signal model for proposed flip-flop is developed and analyzed. Finally, a low power circuit is designed according to the analysis conditions of small signal model and the application requirements of GNSS receivers. Experimental results indicate that the proposed frequency divider works from 6.55 to 0.25 GHz, its consume current is only 0.8 mA, and the core area is 0.014 4 mm2. The proposed quadrature 2∶1 frequency divider is implemented in a 0.13 μm CMOS process, and can steadily work in a sub 1 V supply voltage. It has been successfully applied to low power portable GNSS receivers.

To improve the time synchronization precision of a three-line stereo mapping camera, a time synchronization error model for the three-line stereo mapping camera was set up based on the analysis of working principle of the time system in the camera. Then, several methods such as reduction of row synchronization counter polling period were adopted to optimize the time system. As the time synchronization errors may change with the working time sequence of the camera controller, a real-time time synchronization precision measuring method for the three-line stereo mapping camera was put forward to measure and record the dynamic changes of time synchronization errors in real-time during photographing process globally. With the method, the difficult problem that custom instruments could not be used for long time measurement of time synchronization errors because of their finite memory depths was solved. This method was used to measure time synchronization precision of the time system for the optimized three-line stereo mapping camera. Experimental results indicate that time synchronization errors of the camera are no more than 74.8 μs during photographing process and the probability of absolute value of time synchronization errors no more than 11.2 μs is more than 0.95, which meets the demand of qualification.

The attitude tracking control of staring imaging of an agile small satellite driven by Double-Gimbaled Control Moment Gyroscopes (DGCMG) was investigated. Firstly, an actuator configuration scheme was developed according to the characteristics of agile satellite and the staring imaging mission requirements. Then, the orbit information was used to compute the relative attitude angle and attitude angular velocity of a staring imaging target with respect to the satellite body reference frame, and the proper control law and steering law were introduced to avoid the singularity of Control Moment Gyroscope(CMG). Finally, based on the “SY-3” satellite attitude and orbit control simulation platform, the agile small satellite closed-loop control simulation system was established by taking the mission requirements and the actuator adjustments into account and the proposed design scheme and control method were verified in a mathematical simulation. Obtained results show that the presented solution is effective for staring imaging attitude maneuver control of the agile small satellite and the least gimbal angle rate of DGCMG is the most important factor to determinate attitude tracking precision.

According to the imaging gaps caused by scrolling imaging in the conventional assembled focal plane for a push-broom camera, a mechanical assembling method with uneven overlapping pixels was proposed. Firstly, a model was established to simulate the imaging process when the camera was scrolled. Then, the reason that the gaps appear was analyzed, and a mathematical formula for the gaps between the CCD viewing fields at a certain scrolling angle was derived. Accordingly, the allowable minimum quantity of the overlapping pixels between every two CCDs was derived as well. On these bases, a function of overlapping pixels of images was given for subsequent software programs to eliminate the superfluous pixels and to realize the gapless assembly. Analysis indicates that the error ε of this assembling method is less than one pixel, which satisfies |ε/D|<1. Finally, an example was given.

A new type of 3-DOF sun parallel tracker was designed to track the sun all-around and to maximize its effective output powers. By using the cardan joint on a triangular platform to transfer the most of the weights of solar panels to the stand and to change the positional posture of the solar panels in all direction through three thin steel ropes, the track mechanism could reduce the electrical power consumption of own motor and could generate more efficient powers. Firstly, the forward position equations of the 3-DOF parallel tracker were established based on the theory of coordinate transformation. And then the forward position equations were solved by Newton's iterative method numerically. Finally, the space positional posture was measured and analyzed with the theory of the space parallel projection to verify the accuracy of the forward position equations. A driven energy experiment was performed to contrast the power consumption between the designed mechanism and a traditional two-axis tracker. The experimental results show that the established forward position equations and measured results have the same trend basically, and the average errors of output positional postures for space angles α, β, and distance zB are 1.8%, 2.6%, 0.84%, respectively. This accuracy can meet the requirements of the tracking mechanism for errors. Moreover, the power consumption of the designed tracking mechanism is about 25% that of the traditional two-axis tracker.

As the positioning precision of piezo-actuators is always severely deteriorated by hysteresis nonlinear effect, this paper proposes a neural network control scheme with a hysteresis compensator based on sliding-mode technique to improve the performance of the piezo-actuators. A Radial Basic Function Neural Network (RBFNN) was developed as a equivalent control value in the sliding-mode control and the hysteresis compensator was used to estimate the lumped uncertainty caused by the varying parameters in the RBFNN, external disturbance and the approximate algorithm to compensate the output signal of the RBFNN. For the above steps, the dynamics of actuator was guaranteed on the sliding surface. The adaptive tuning laws of the network and the compensator were derived on the basis of Lyapunov stability theory, and the convergence and stability of the control system were proved theoretically. A low frequency triangle reference displacement with a variable amplitude was used to detect and analyze the effect of the proposed control method. Experimental results show that the mean and maximal positioning errors by the tradition neural network are 0.43 μm and 0.77 μm respectively, but these errors can be reduced to 0.27 μm and 0.49 μm under the sliding model controller. Finally, the positioning precision is approved evidently.

The cryogenic hysteresis and creep characteristics of piezoelectric bimorph scanners for cryogenic optical systems in space infrared cameras were studied. The piezoelectric bimorph has superior characteristics of large displacement, high resolution, no self-heating, etc., moreover, it can work efficiently at a cryogenic temperature. However,the inherent hysteresis and creep characteristics of the piezoelectric bmorph have a large impact on the scanning accuracy when it is used for high precision scanning. In this paper, the hysteresis and creep characteristics of a fabricated piezoelectric bimorph scanner were experimentally studied at room temperature and cryogenic temperature and the results were compared. The results indicate that both hysteresis displacement and creep displacement of the piezoelectric bimorph scanner have a large decrease, but the decrease of total travel displacement is much larger, which leads the hysteresis rate and creep coefficient increase more large. The hysteresis rate at 120 K is twice that at 300 K and the creep coefficient at 120 K increases by an order of magnitude as compared with that at 300 K. Results mean that the hysteresis and creep characteristics of piezoelectric bimorph scanner at cryogenic temperature are more serious than it at room temperature, which will bring greater impact when it is used for high precision applications.

To improve the accuracy and efficiency of the simulation for the optical system in a Micro-electro-mechanical System (MEMS) and to overcome the difficulties in the system-level modeling of MEMS optical components, a system-level modeling method was proposed to support the co-simulation for MEMS system-level mechanical components and electrical components. First, the modeling methodology of Multi-Port-Element Network (MuPEN), the characteristics of Gaussian beam and the theory of space coordinate transformation were introduced. Then, the modeling process of a micro-mirror was given to explain the modeling method of all optical components. Finally, using the hardware description language of Verilog-A, an optical library including some typical components was established. The non-differential voltage simulation results of the system-level scanning system were compared with those of CoventorWare and results show that the scanning system in former environment can eliminate the blind area, and the maximum error is within 3%. It indicates that the proposed modeling method for optical components could work effectively and accurately, and the study is significant value to the MEMS system-level design.

A low-g micro inertial switch based on the planar rectangular helical spring was developed by utilizing a SOI (Silicon-On-Insulator) wafer with double buried layers and the manufacturing technology including KOH etching, Inductance Coupled Plasma (ICP) etching, anodic bonding and spray coating. Based on the self-stop technique for KOH etching and ICP etching of the buried layers, the precision of the spring thickness was controlled to be ±0.46 μm. The electrical property of the SOI wafer was analyzed. The anodic bonding between borosilicate glass and SOI wafer was successfully carried out utilizing the uniform electric potential technique,and an anti-stiction structure with a size about 200 μm×200 μm was fabricated using the maskless pyrex-etching in the process of wafer cleaning and drying. Finally, a silicon-base was used to solve the Au-Si eutectic problem resulted from high temperature during ICP etching process. Experiments show that the chip yield is greatly increased by optimizing the fabrication process and the fabrication process can support the production of low-g micro inertial switches in batches.

Three-dimensional scene is the mirror image of recording scene in a traditional integral imaging system. However, when a camera array is employed to capture elemental images, it can not be reconstructed correctly becaused of the mismatch between the parameters of lens array and the parameters of camera array. In order to rebuild a new relationship between the reconstructed image distance and the parameters of elemental images, a new method by using the Space of Corresponding Image Points (SCIP) to calculate the reconstructed image distance is proposed. In this method, the SCIP is a function of the object distance of scene, the focal length of camera and other parameters. It is equivalent to the lens array in the process of image reconstruction. By transforming the SCIP matched with the parameters of lens array, the scene with different reconstracted image distances can be reconstructed exactly. By experiments, this paper verifies that different SCIPs can result in different reconstructed image distances theoretically and experimentally.

For the unreliable data storage problem caused by bad blocks and single event upsets for the NAND flash memory in a space camera, this paper explores a bad block management strategy and an error correction algorithm. Firstly, the bad block management strategy based on parallel double-traverse mechanism was proposed by analyzing the characteristics of structure and operation for the NAND flash memory, the design ideas of the double traverse mechanism were described and its effectiveness was analyzed. Then, the error correction algorithm based on the shortened code RS (246,240) and RS (134,128) in the field GF (28) was proposed, and the encode/decode algorithm and corresponding circuits were given. Finally, the verification experiments on an image storage platform in the prototype machine for a space multi-spectral camera were carried out. The experimental results show that the bad block management strategy can fast and reliably dispose the bad block events, and the algorithm can identify the bad blocks in one system clock period. The error correction algorithm can correct 27 B error within the 2 KB/page with a encoder speed of 72.53 MBps and a decoder speed of 54.26 MBps. Proposed stratege effectively solve the problem of unreliable recording data in the NAND flash memory.

As existing filtering methods could not overcome the contradiction between noise removal and feature preserving for the light-spot signals on a CCD, the random error resulted from image noises has become a key constraint on the location accuracy of the light-spot images. Therefore, a filtering method named Attribute Distance Weighted Average (ADWA) is proposed to denoise the light-spot images. The method separates signals and noises based on their attributes, and alleviates the contradiction between noise removal and feature preserving of the signals progressively through introducing new attributes. Experimental results show that as compared with the bivariate ADWA based on “location” and “value”(2-ADWA/LV),the sub-pixel location accuracy of light-spot has been improved by more than 20% and the standard deviation of location reduced to 0.024 pixel by introducing a 3-ADWA/LVG formed by adding an attribute "gradient" to 2-ADWA/LV. It concludes that when the ADWA is used to denoise the light-spot image, the filtering performance can be further enhanced progressively in both noise removal and feature preserving through introducing new attributes. The ADWA gives a new effective way to further improve the sub-pixel location accuracy of the light-spot.

To avoid the damage of Time Delay and Integration(TDI) CCD sensors in the experimental process, a TDICCD simulation instrument synchronized with input driving clocks is introduced. The simulation instrument is designed with a simple method to meet its physical size in a length of 100 mm and a width of 45 mm. As the size is same as that of a real TDI CCD sensor, the instrument can simulate the photonic characteristics of the pixels for TDI CCD sensors exactly. Furthermore, it can simulate the electrical characteristics of pins in the TDI CCD sensor and can synthesize a complete video output signal, by which an important function, selectable TDI stages with 16,32,48 or 64 stage, is available. Moreover, the instrument can be taken to simulate the thermal delay characteristic between the input driving clock and output OS signals in the temperature adjustable range of 0.5~12.5 ns, which is convenient to the sample time adjustment for the signal circuit.

An image registration method based on complex Zernike moment phase angle estimation was proposed. Firstly, the Harris-laplace operator was used to detect interest points in an image, and the interest points were regarded as initial feature points. The Zernike moments defined on the scale normalized interest point neighborhood were computed, and a new robust estimation method for phases was presented to compute the rotation angle between two normalized regions. Then, the magnitude and phase angle information of Zernike moments were combined and used to measure the Euclidean distance between two matching regions. Finally, an iterative refined angle method was proposed to estimate the parameters accurately, and the image registration was finished after the geometric transform of input images. The experimental results show that the proposed algorithm impletments a precise image registration under the scaling, arbitrary rotation and noise. The average coverage percentage achieves 94.125% when the rotation angle error is less than 20°, which reduces the false match rate effectively.

To modify the geocentric core distance error of a Time Delay and Integration (TDI) CCD space camera in the calculation of image motion compensation, and decrease the influence of the error on the match error of image motion velocity, a model of image motion calculation of points bellow satellites was deduced. The influence of the geocentric core distance error on the relative error of image motion velocity was analyzed based on the model, and the geocentric core distance error was modified by two steps according to the error sources. The former adopted the World Geodetic System (WGS-84) to modify the error generated by ellipticity of the error, and the latter used the data source of the earths height above sea level (United States Geological Survey Digital Elevation Model, USGS DEM) to make a digital elevation map modify the error generated by the deference heights above sea level. After modifying, the image motion model of points bellow satellites was deduced. The calculation and analysis by the model indicate that the maximum error of image motion velocity eliminated are about 2.85% and 1.76% with the modification of geocentric core distance error by the WGS-84 model and digital elevation map,respectively. Those data clearly show that the match error of forward direction(the integral direction of TDI CCD)has been greatly decreased by modifying the geocentric core distance error, and the imaging quality of TDI CCD space camera has improved greatly.

An automatic cavity adjustment system was established to improve the quality of laser gyro cavity adjustment and to overcome the drawbacks of manual detection such as low efficiency and low anti-interference capacity. According to the process of the cavity adjustment, a multi-sensor information fusion architecture was established by a CCD camera and a photomultiplier. A detection method of laser gyro cavity adjustment was proposed based on the D-S evidence theory. By analysis and calculation of the signals detected by the CCD camera and the photomultiplier, the center coordinate difference between facula and diaphragm and the loss value of the laser gyro were obtained and the evaluation function for each cavity adjustment result was deduced by these data. Furthermore, the qualified and unqualified belief functions of the cavity adjustment were obtained, respectively, and the quality of the cavity adjustment was verified based on the maximum support rule. The experimental result indicates that the accuracy of the detection method based on D-S evidence theory is 91.14%, which improves the quality of cavity adjustment effectively and validates the feasibility of the proposed method.

As an ideal 3D display way， holographic 3D display can present the same information of depth and parallax as real objects. However, the computer-generated hologram for a 3D object has to do complex and massive computations, how to create a computer-generated hologram quickly for the 3D object has become a key problem in the digital holographic 3D display technique. In this paper, the key techniques of digital holographic 3D display are discussed firstly, including three kinds of methods for creating computer-generated holograms like object point scattering, holographic stereograms and holographic tomography, a RGB separating method for colorful holographic 3D display, and several kinds of improving quality methods for reconstructed images. Then, few latest holographic 3D display systems and key devices are analyzed technically. Finally, it summarizes the present state of the digital holographic 3D display technique and points out that the technology will develop to real-time, dynamics, larger sizes and higher resolution.