As the imaging performance of an optical system is related to its transversal magnification, this paper proposes a transversal magnification measurement scheme based on wavefront error metrology for the optical system. On the basis of that the locations of image plane and object plane satisfy the Newtons and Gausss formulas in a given optical system, the defocus in wavefront error was used to monitor the movement of image point. According to the minor movement of object point, the longitudinal magnification could be calculated by the differential of Newtons and Gausss formulas, and the transversal magnification was obtained finally. In order to validate the scheme, a model of transversal magnification measurement was established, and the criterions of the object point shift and the initial defocus were summarized. Then, the influences of geometric tolerance and positioning accuracy of the image point on transversal magnification were analyzed. Finally, an experimental platform based on a point diffraction interferometer was established for measuring the wavefront error of a reduced projection system, and the transversal magnification of the system was measured. It shows that the difference between theoretical and measured results is lower than 0.24%, which verifies the feasibility and veracity of this scheme.

To realize a broad spectrum and dynamic local high-resolution imaging, an off-axis reflective active zoom system is designed. This system combines off-axis reflective active zoom theory and local high-resolution theory together and realizes dynamic variable resolution in different focal lengths. Deformable mirrors(DMs) with variable curvature radii are used to change focal length to avoid complicated mechanical zoom movement control in a traditional system to reduce the volume and weight and to guarantee the broad spectrum of optical system. Through the control of deformable mirrors at different focal lengths, high-resolution within the Region of Interest(ROI) is achieved and the amount of data transmission is reduced. Moreover, as reflective systems have no chromatic aberration, it overcomes the disadvantages of monochrome imaging of traditional local high-resolution systems. After optimization, the imaging spectrum is visible light, the focal length ranges from 75 mm(Field of View(FOV) x: 0°—0.5°, y: 3°—10°)—150 mm(FOV x: 0°—0.5°, y: 1.7°—5°), the F/# is 7—14. The result of theoretical and simulation analysis indicates that the imaging quality of the ROI reaches the diffraction limit, and the system realizes dynamic local high-resolution in the FOV.

A 2D temperature field scanning measurement method in a combustion flow based on Coherent Anti-stokes Raman Scattering (CARS) was explored for temperature distribution measurements. By synchronized scanning incidence lasers and exit CARS signals, the temperatures of discrete points for methane/air premixed flame in a horizontal section were measured and a 2D temperature field was reconstructed by using 2D linear interpolation. The temperatures of the flame in the same space position in scanning processing was measured with synchronizing scanning method, by which the additional effect of flame space distribution in homogeneity on scanning measuring temperature was removed. Finally, the measuring uncertainty of the scanning CARS system was analyzed in a given experimental state. The experimental results for scanning measuring the same space position show the average temperature in the point is 2 074 K, meanwhile, the measuring uncertainty of the scanning CARS system is accounted below 21 K at the same experimental condition. In conclusion, the research quantizes the measuring uncertainty of the scanning CARS system, improves the reliability of results of scanning temperature measurement and provides an experimental foundation for the high precision measurement of flame temperature distribution, simulation validation of Computational Fluid Dynamics(CFD) and the investigation of flame problems.

Existing evaluation and testing methods are difficult to apply to the wavefront aberration test at different elevation angles for a photoelectric telescope with 3—4 m aperture. Therefore, this paper presents a Wavefront Error(WFE) test method by subaperture wavefront slope discrete sampling and then reconstructing the full aperture wavefront aberration. On the basis of the co-simulation by mathematical analysis and optical simulation, it studies the relationship between wavefront reconstruct accuracy and sub-aperture scanning motioned tilt error, sub-aperture scanning position tilt error, image point coordinate measuring error and wavefront reconstruction algorithm uncertainty. Collaborative simulation results show that the relative error(ΔPV) of iterative algorithm is about -0.002 8λ (λ=632.8 nm), and the relative error(ΔPV) of pattern algorithm is -0.002 7λ. When the sub-aperture tilt error is less than 0.2″, the wavefront reconstruction error(ΔPV) is about 0.02λ. When the sub-sampling aperture position error is less than 0.2 mm, the wavefront reconstruction error(PV) introduced is less than 0.04 nm. And when position error of imaging point is less than 5 μm, the wavefront reconstruction error (ΔPV) introduced is less than 0.03λ.The results demonstrate that the pattern algorithm has a higher error margin, and the convergence is better when the measurement errors are considered. Moreover, it suggests that it should take the angle monitoring and error compensation mechanisms into account and the angle measuring accuracy of the sub-aperture tilt monitoring system should be better than 0.2″when an actual measuring equipment is established.

A high-speed particle image velocimetry based on 90° Mie scattering was proposed to evaluate the status of laminar flow and single-cell flow stability in the flow cell of a flow cytometer. The velocity stability of the particles in the flow cell was used to evaluate the single-cell flow stability of the flow cytometer. Firstly, 90° Mie scattering images of the particle were acquired with a high-speed image sampling system. The 90° Mie scattering selected could avoid the interference of the excitation light source and could remove the background source and improve the image contrast. Then, the grey cluster analysis algorithm based on trapezoid whitenization weight functions was used to classify for all different types of images with shortages, normal, diffraction and overlap. Finally, a center-point method was taken to determine the boundaries of the trailing image to improve the computational accuracy of the length of particle trailing. A experimental system was carried out to verify the feasibility of the comprehensive method. The results show that the method obtains distinct trailing images and they have been classified accurately. Moreover, for the system of this article, the average length of trailing image covers 116.9 pixels, and the standard deviation is 1.7.

According to the demands of Chip-on-board(COB) wafer level packing Light Emitting Diode(LED) powder complement and discharge equipment for the photoelectric parameter measurement, this paper develops a self-designed fast LED photoelectric parameter measurement system based on a fiber optic spectrometer. The system consists of an optical parameter detection module, a mechanical structure of LED measurement and a display module. The optical parameter detection module mainly consists of a homemade spectrometer for acquiring the spectral data to get the chromatic parameters of the LED. The mechanical structure of LED measurement consists of an integrating sphere and a test platform to be mounted different holders of COB packaging LED. It can measure the luminous flux, color coordinates and color temperatures fast. The designed LED photoelectric parameter measurement system is used to scan the COB packaging LED and measure photoelectric parameters, meanwhile the operator can repair powders based on actual measurement results. The results for the measurement of 10 LEDs show that the single measurement time is less than 3 s, the color coordinate accuracy and color coordinate repeatability are better than 0.003 and less than 0.000 5, respectively, and the color temperature precision and color temperature repeatability are 0.6%@5700K and less than 0.000 8, respectively. It satisfies the requirements of measurement system for the high-power COB packaging LED and is characterized by fast speeds, higher accuracy and higher repeatability.

When a thin-film filter is used in oblique incidence, the central wavelengths of the polarization light will be separated obviously, and it will cause serious polarization dependent loss. Therefore, this paper designs a 100 GHz channel spacing stack to depolarize in the central wavelength based on the equivalent layer theory and to realize the angle and wavelength tunes of the thin film filter. Firstly, the depolarization equivalent refractive index in the space layer of the filter was calculated by a phase analysis and to complement the alignment of the different polarization lights. Based on the equivalent layer theory, a symmetric three layer stack was designed to replace the depolarization equivalent refractive index of the spacer. As comparing with the original five layer low polarization thin film filter, the three layer spacer stack is simpler, and it is more accurate to replace the depolarization equivalent refractive index. The simulation and experimental results indicate that the stack can align the polarization light central wavelength from 0° to 20°, in which the polarization light separation is less than 0.03 nm and the wavelength tuning range reaches 35 nm.

To achieve the auto control of transition observation of the Optical Lattice Clock transition developed by the National Time Service Center, Chinese Academy of Sciences, a complete control system is designed. The control system consists of a timing sequence control unit with delay accuracy and synchronizing precision in μs level and a laser spectral scanning control unit. The two units were both realized through the virtual instrument exploited by LabVIEW software programming to control the optical field and magnetic field precisely. The two-level cooling and optical lattice trapping of strontium atoms are implemented and the high signal-to-noise ratio(SNR) clock transition 1S0-3P0 spectral line of strontium atom with a line-width of 180 Hz is obtained. The spectral line with higher SNR and narrower line width indicate that the optical lattice clock can operate in higher stability and whole control system meets the precision requirement of the running of strontium optical clock system. Moreover, the experiment demonstrates that the control system has universality, and can be extended to other systems to control the optical fields and magnetic fields.

The measuring methods for normal emissivities of materials at -60—50 ℃ was explored. To achieve accurate measurement of normal emissivity of materials at temperature range of -60 ℃ to 50 ℃, a measurement model of normal emissivity of materials was established based on the emissivity definition. To shield the effects of environment stray radiation and atmospheric absorption, a measurement facility of the materials at low temperature was built by utilizing a vacuum and liquid nitrogen background channel. The normal emissivities of two samples (copper oxide and high emissivity ceramic) were measured by this facility. The results show that normal spectral emissivities of two samples decrease with wavelengths. Moreover, with temperature increasing, the normal integral emissivity of the copper oxide keeps stable at the range of 0.850±0.012 and that of the high emissivity ceramic reduces by 0.124. Finally, the infrared spectral irradiation was acquired in higher precision at a lower temperature condition, the measurement uncertainty of normal spectral emissivities of materials at low temperature was analyzed. The results show that the relative expanded uncertainty is less than 6.0%.

This paper introduces the accurate pointing and auto-unlocking sun tracker on the Fengyun-3(FY-3, C)satellite designed by ourselves. The auto-unlocking sun tracker is rotating around the biaxial tracking platform, consisting of a drive unit, an encoder, a sun sensor, mechanical positioning elements etc. It realizes the on-orbit pointing strategy by both active routine and passive redundancy. In order to avoid the influences of vibration and shock in transporting and launching the satellite on the performance and lifetime of the satellite, the pointing device has a degree of freedom for locking and releasing function, and achieves the function that locking freedom before the launch and releasing freedom after the launch. The paper describes the structures of the pointing device, the design process of precise pointing and the result of life test. FY-3 (C) was successfully launched in October 2013. the sun tracker of solar radiation monitor has unlocked successfully when it was into the orbit. It has run continuously for 17 months in the orbit, and showing the orbit pointing accuracy within ±0.05 °. The tracker loaded the solar radiation monitor has accessed a large number of high precision and valuable solar radiation data. The accurate pointing and auto-unlocking sun tracker provides foundation for FY system satellites and other satellite loads in design of pointing devices.

A system for the fabrication of RFID(Radio Frequency Identification) tag antenna was setup based on a digital drop-on-demand technology. Firstly, the conductive ink was jetted on the polyethylene terephthalate (PET) substrate with a hydrophobic thin film to form the designed antenna pattern. Then the conductive pattern was sintered to form the RFID tag antenna in an oven. The influences of the system parameters on the generation of droplets and the fabrication parameters on the formation of the conductive line were researched. The conductive line with a width of 100 μm,a thickness of 2.8 μm and the resistivity of 5.2 μΩ·cm was obtained, and its resistance was almost proportional to the length. Furthermore, a folded dipole antenna was designed, simulated and fabricated, and the performance of the antenna was measured. The results show that the resonant frequency, bandwidth and the S11 parameters of the antenna fabricated are in agreement with the simulation results. The experimental results indicate that the fabrication of the RFID tag antenna based on the drop-on-demand technology has advantages in simple system construction, low cost, accurate droplet ejection and a flexible antenna pattern. Moreover, the folded dipole antenna is characterized by size controllability, high conductivity, good resistance uniformity and good antenna performance.

To improve the machining efficiency of fused silica optical elements with large diameters, this paper presents a method for heat-assisted ductile ultra-precision grinding of the fused silica. The mechanism of heat-assisted ductile grinding for the fused silica was analyzed, and the effect of different grinding depths on the highest temperature rise on the surface of the grinding zone was investigated by theoretical inference. The fused silica was ground in dry method by a high temperature resistant ceramic bond Cubic Boron Nitride(CBN) wheel and the high efficiency ductile grinding of fused silica was realized by using grinding heat to improve the mechanical properties of fused silica in the grinding zone. Through the grinding experiment, the effect of different grinding depths on the surface roughness(Ra) and sub-surface damage depth of fused silica was investigated. The experiment results show that the Ra and sub-surface damage depth are reduced with the increase of grinding depth. When the grinding depth is greater than the crack depth(5 μm) by coarse grinding, smooth and no crack surface of ductile grinding with the Ra 0.07 μm can be achieved. The mechanism of wheel wear was studied by a scanning electron microscope. The results show that the wear mechanism of ceramic bond CBN wheel ductile grinding of the fused silica is grain flatted, which provides the basis for researching a new type of ceramic bond CBN grinding wheel for dry grinding of the fused silica.

In combination of laser etching and numerical control (NC), a localized and precision removal technology for metal coating on engineering plastics was proposed based on numerical control (NC) laser milling. As different positions of the parts require different qualities, a precise processing technology was presented through study and analysis of laser ablation experiments to guarantee the quality of figure edge and corresponding technological parameters were determined. Then, the removal technology for the metal coating with a constant depth was developed by adaptive controlling the laser fluence on the basis of the practical feed speed, by which the allowance within the pattern was removed and the over-ablation problem of the processed materials caused by that the machine feed speed can not reach to a preset value was solved. Finally, a compound three-dimension signal transmission/receiver was taken as an example, the feasibility of the proposed method was verified by processing the part pattern regionally. Processing results show that the compound three-dimension signal transmission/receiver has been processed precisely with the localized and precision removal technology based on numerical control (NC) laser milling. The join of the plane part and the conical part is accurate, the neat micro-strip edge is smooth and almost no heat affected zone. The results meet the requirements of manufacturing the part with high quality and high efficiency.

The compliance features of flexure hinges were analyzed to design and optimize the flexible amplification mechanisms. A new general structural parameter called compliance ratio λ was proposed. The sensitivities of main type of output displacement of the flexure hinges with different compliance ratios λ were analyzed, and the effect laws of compliance features on the commonly used flexure hinges were also discussed in detail. Then, by taking the compliance ratio λ as the basic parameter, the theoretical calculation method of the amplification ratio of a two-stage lever-type flexure hinge mechanism was proposed in considering the offset of the flexure hinges rotation. The optimization design of the flexible amplification mechanism was studied on the basis of the characteristics of the compliance ratio λ. The finite element simulation and the experiment were performed. The results show that the amplification ratio of the optimized flexible amplification mechanism is larger than that of the flexible mechanism before optimization, and the increases of the simulation and the experiment are 0.234 and 0.23, respectively. These results demonstrate that to optimize the flexible amplification mechanism based on the compliance ratio λ of flexure hinges improves the amplification ratio and the working range significantly, and enhances the accuracy of movement and positioning of the amplification mechanism.

The structural software ANSYS and the fluid software ANSYS CFX. were used to simulate the structural-fluid coupling of a valveless piezoelectric pump and to research its output performance. Three different kinds of valveless pumps, the pump with middle inlet, the pump with middle outlet and the pump with side inlet/outlet were designed. The structural-fluid coupling on these valveless piezoelectric pumps was simulated by the software ANSYS and the software ANSYS CFX. The simulation results show that the pump with middle outlet has the maximum output flow rate. The prototypes of the three kinds of piezoelectric pumps were fabricated and a corresponding testing system for the pumps was set up and test experiments were performed under a sinusoidal AC excitation signal of 45 V and the frequency of 0-700 Hz. The results show that the maximum output flow rates of the three kinds of valveless pumps are 3.8 ml/min, 6.0 ml/min and 4.0 ml/min, respectively. Among them the pump with middle outlet has the maximum output flow rate, which is consistent with the results of the structural-fluid coupling simulation and verifies that the simulation method proposed by this paper can guide the design of piezoelectric pumps.

To increase the axial stiffness of a Lamina Emergent Torsion( LET) under the condition of invariable rotational stiffness, a new tensile flexure hinge was designed by improving traditional LET structures. Base on the structure of Lamina Emergent Mechanisms(LEMs), the whole LET structure was equal to a spring stiffness model. By modeling the spring stiffness model in theory, the closed-form solution was obtained. Then, a Finite Element Analysis ( FEA) model was set up by the ANSYS to analyze the deformations under the rotating load and axial load and to compare with the previous theoretical model. Results show that the equivalent stiffness solution based on spring stiffness model is consistent with that of the simulation analysis, in which the bending tensile stiffness of the tensile LET is only 1.2 times that of the LET, but tensile stiffness is 76.43 times that of the LET. It indicates that the bending stiffness does not increase obviously, but the tensile stiffness of the tensile LET has significantly increased effectively and the tensile capacity of the LET is improved greatly. The design of the tensile LET meets expectation.

A compact Fast Steering Mirror(FSM) with a rigid support structure was designed for adapting to the great vibrancy, impact and high-low temperature conditions of an airborne platform. The support shafting, actuators, angle measurement elements of the FSM were designed and selected respectively on the basis of design requirements of the FSM. Then, a device to adjust shafting clearance was designed to improve shafting precision and provide additional supporting for mobile parts of the FSM. The special grating displacement sensors with small sizes were designed and four grating sensors were placed on the diagonal symmetrically to measure the position of the mirror by quadratic average of each measuring result, which reduces the volume of the FSM, improves its measuring precision, and removes the influence of clearance in the shaft direction on measuring results. Finally, the control bandwidth and pointing precision of the FSM for the airborne platform were tested, and the results show that the designed FSM offers the control bandwidth about 110 Hz, azimuth pointing error less than 3.4″ and the pitch pointing error less than 3.8″, which satisfies application requirements of vehicle platforms.

This paper focuses on finding out a optimum quadrature error correction method for a dual-mass Micro-electro-mechanical System (MEMS) gyroscope with quadrature error correction combs and sense feedback combs. The structure of the gyroscope was introduced, and influences of quadrature errors on the output signals were simulated and quantized. The results show that quadrature error equivalent angular rate is 200(°)/s and the demodulation phase error is less than 2°, which generates 15(°)/s variation of output signal. Three kinds of quadrature error correction methods used commonly in measuring this structure were investigated, and they are Charge Injecting Method (CIM), Quadrature Force Correction Methods (QFCM) and Quadrature Coupling Stiffness Correction Method (QCSCM). The feasibility of the three method was verified in theory. A gyroscope without quadrature error correction was tested, the results show that the quadrature signal peak-peak amplitudes from left and right mass preamplifiers are 150 mV and 300 mV respectively. Then, the correction method based on left mass and right mass independently was proposed. The the experiments based on CIM, QFCM and QCSCM were performed. The results show that the quadrature error signals are eliminated basically, their bias values and bias stabilities are -8.361(°)/s and 423(°)/h, 2.419(°)/s and 82(°)/h, 1.751(°)/s and 25(°)/h, respectively. According to the results, the QCSCM is verified to be the optimal method for quadrature error correction, and the theory analysis result is correct.

An on-line linearity calibration platform for capacitive displacement sensors is proposed to implement the high-precision adjustment and the measurement of displacement. The symmetry axis for movement, the measuring axis of an interferometer and the measuring axis of a sensor are collinear in the platform, so that the Abbe error is decreased in principle. For the z/tip/tilt adjustment function in the platform, the alignment between the sensor and the target surface is realized. The composition and principle of the calibration method are introduced and the micro-displacement is adjusted by a symmetrical parallelogram mechanism. Then, the output compliance and stroke of the guiding mechanism are analyzed based on Compliance Matrix Method(CMM). The experiment result demonstrates that the stroke of the calibration platform is 735.162 μm and the errors are 7.410% and 4.633% comparing with that of the Finite Element Method(FEM) and CMM, respectively, which meet the requirement of the stroke. Moreover, the sensor linearity is improved from 0.014 21% to 0.006 231% after calibration calculation. The linearity calibration method has high-precision and it satisfies the requirement of fine displacement adjustment of the mechanism.

To meet the requirements of the Tertiary Mirror System (M3S) in a Thirty Meter Telescope(TMT) for mass and stiffness, how to allocate rationally the stiffness for different parts in the system was researched. As the first-order resonant frequency of M3S should not be less than 15 Hz, the composition of M3S was researched, then a four-point lumped-mass model was established by the pre-designed data. Furthermore, the stiffness of the mirror support system was calculated, and an eigenvalue inverse solution was used to obtain the stiffness matrix of a simplified model in six global directions. Finally, stiffness of all springs in the system was given, and the simplified model and the calculated results were used to guide the structure design and control design later. The simulation software Adams was used to verify the stiffness allocation and analysis process. The results show that the stiffness allocation of the M3S is reasonable and the first modal frequency of the system can achieve to 15.2 Hz, which meets the design requirements. With the stiffness allocation, the system design can offer higher efficiency and rationality.

To achieve the decoupling between the lifting movement and the horizontal movement of a forging manipulator, a kind of manipulator mechanism to achieve the decoupling was presented based on the structural composition principle, and the motion mechanism and the dynamic performance were studied and verified by experiments. Firstly, the composition principle of a fully decoupled forging manipulator was introduced based on the Hoekens four-bar mechanism. The degree of freedom of the fully decoupled forging manipulator was analyzed by using the screw theory, and the decoupling mechanism between the lifting movement and horizontal movement was also dissected. Then the dynamic model was established by using the principle of virtual work. The mechanical behavior between the new mechanism of forging manipulator and the typical parallel-link one was compared and analyzed in the lifting movement and the forging process. Finally, an experimental model of the new forging manipulator mechanism with ratio 1/20 was developed to verify its decoupling performance. Experimental results indicate that during the lifting movement, the displacement variation of the gripper carrier in the horizontal direction remains within 1 mm, which basically verifies that the fully decoupled forging manipulator mechanism achieves the decoupling between the lifting movement and the horizontal movement.

The main factors effect on isolation control accuracy of a gyro-stabilized platform were analyzed, including the un-modeled unit in a controlled system, the random disturbance of state and the measurement noise of an output signal. The control schemes to overcome these effect factors and improve the isolation accuracy of a gyro-stabilized platform system were explored. In order to improve the control accuracy, an integrated solution for eliminating all effect factors was researched and a two-step control strategy was proposed. The first step is to employ the Active Disturbance Rejection Control (ADRC) to observe and compensate the un-modeled unit and to design the feedback control in ADRC as PID controller to control the compensated system. The second step is to use a Kalman filter to eliminate the random disturbance and measurement noise. The control scheme was described in details and its performance was simulated. The results indicate that the isolation degree reaches 4.61% by using this control strategy when the disturbance of the platform is 3° and 1/6 Hz, which means the isolation performance has improved by 50.9% as comparing with the performance 9.39% from the control strategy which consists of the parameter identification of nonlinear friction and its forward compensation. The control strategy evidently has a higher practical application value.

The attacking orbits of interception satellites were explored when target satellites flied with several small satellites in concomitant formation in anti-satellite combination. By taking an interception satellite with a low continuous thrust to transfer orbit for a researching object, an attacking orbit optimization method based on genetic algorithm was proposed. A dynamic defending model of the small formation satellite was taken as environment model, and comprehensive fitness function was constructed according to the requirements of orbit safety and fuel saving. Then, the coding method, selection operator, crossover operator and the mutation operator were designed for proposed algorithm. A simulation experiment was performed on the MATLAB platform. The simulation results show that the interception satellite hits the target satellite at 650 s, spending 410 s. The algorithm proposed searches an optimal attacking path and has a good convergence and stability. As compared with the same kind of optimization method, this algorithm effectively reduces the boot time of rockets, and reduces the task loads of orbit maneuver attitude control of the interception satellite.

When traditional Bang-Bang control in optoelectronic equipment is used in fine position of double-targets, the system is difficult to be stabilized because of its amplifying coefficient in infinity. This paper analyzes shortages of the Bang-Bang control, and proposes a kind of variable structure control, namely the coarse tracking algorithm based on the Bang-Bang control. It uses Bang-Bang control in the coarse tracking and switches it into linear control in a small deviation range. In the linear control, the strategies of infrared first-order capturing and infrared second-order tracking are proposed depending on characteristics of different regulators, and switching guidelines for the Bang-Bang control and linear control, and the infrared first-order capturing and infrared second-order tracking are given. The experiments show the proposed method makes the photoelectric equipment spend around 2.3 s in 90° back and forth position, which is shorter nearly 1 s than that of traditional method. Meanwhile, the laser rangefinder gives the rate of three-dimensional information to be 0.45 Hz, and it improves the optoelectric countmeasuring ability of the system for double targets.

A new method based on the linear model of image gray level was proposed to estimate sub-pixel jitters of a sequence image for cloud scene acquired by staring remote sensing imaging systems. Firstly, the image pixel and corresponding neighborhood gray levels were mathematically described by using linear model with three parameters, and the image gray was modeled. Then, by taking the jitter parameter as optimization variables in the linear mathematic model and the comparability between image sequences and reference frame as optimization objective, a new estimation method for the sub-pixel jitter was proposed based on least square optimization approach. Subsequently, the solving equation was derived. Finally, the method was verified by using simulated image sequences containing the cloud scene. Experimental results indicate that the proposed method is able to implement the subpixel estimation effectively and offers the estimation accuracy no less than 0.1 pixel. The obtained estimation accuracy is higher than that of the conventional feature-based ones, and can be used in geometric calibration and target positioning of remote sensing images as well the multi-frame relative detection of time-series images.

To eliminate the influence of Rolling Shutter(RS) on imaging quality in higher speed imaging by a CMOS (Complementary Metal-Oxide Semiconductor) aerial camera, a mathematical model to calculate the RS effect at arbitrary gesture angles was built . The CMOS imaging principle was analyzed, and each pixel velocity of CMOS array was derived from coordinate transformations. After analysis of the working principle of the RS, the analytical equation of the RS effect was deduced. The Monte Carlo statistical method was used to analyze the model accuracy, the key parameters of the model was simulated and the influence of key parameters including interframe delay and gesture angles on the RS distortion were discussed. The experimental results indicate that when the measuring errors of height, speed and the gesture angles are less than 0.09 km, 0.3 km, and 0.02°, respectively, the calculation error of this model is less than 1/3 pixel. These results demonstrate the effectiveness of the model. It concludes that this model can be the theoretical foundation of quantitatively analyzing RS effect of large frame CMOS aerial cameras and can offer some theoretic guidances for applying CMOS in the field of aerial cameras.

When Support Vector Machines(SVMs) are applied in airborne LiDAR point data classification, their performance is limited by weak model sparseness, the prediction lack of probabilistic sense, and long training time. Therefore, a novel LiDAR point could data classification method was proposed based on an Informative Vector Machine (IVM). Firstly, the assumed density filtering was utilized to produce an approximation for probit classification noise model, and the classification problem was transformed into the regression problem. Then, the informative vectors of the active set in LiDAR point cloud data were chosen to achieve the model sparseness according to the largest posteriori differential entropy. Finally, in the training process, the kernel parameter was obtained by Marginal Likelihood Maximisation(MLM) and an One Against Rest (OAR) classifier was selected to realize multi-class classification. The LiDAR point cloud data from Niagara and Africa were selected for experiments in comparison with the SVM, and experimental results show that the classification accuracy of the method based on IVM increases to 94.20% and 90.78% respectively, the number of basis vectors reduce to 50 and 90 separately, and the training time decreases to 5.86 s and 8.03 s respectively. In conclusion, the classification method based on IVM has advantages in fast training speeds, strong model sparseness and high classification accuracy.

A saliency detection algorithm for infrared images based on texture coarseness was proposed to detect the saliency of targets owing to a low image contrast. Firstly, Tamura's principle of calculating coarseness was researched, and a new method to calculate the coarseness was presented by analysis and evaluation of the coarseness. Then the image was decomposed into a set of super pixels and the maximum mean intensity difference of the super pixels was calculated. Furthermore, the best scale of super pixels was defined by using maximum mean intensity difference to be a measure of the texture coarseness. Finally, the region of every super pixel was expanded isotropically and the saliency of infrared image was measured based on the local contrast and grey information of the texture coarseness with the weight of intensity. The effectiveness of algorithm was verified. Results show that coarseness based on the proposed method remains unchanged under a noise level of 10% and the homogeneity is better in the feature map of coarseness. Meanwhile, there are many miscellaneous points in Tamura's feature map of coarseness. Compared with other methods of saliency detection for infrared images, the proposed algorithm has the highest hit rate, up to 0.752. The algorithm exploits the feature of texture coarseness, and provides a new selection method for the saliency detection of infrared images.

The traditional Crossmedia Relevance Model(CMRM) is based on the relevance between visual information and annotation words, while ignoring the inter-word semantic relevance. Therefore, a new CMRM image semantic annotation model based on a graph learning was proposed. Firstly, the ontology of a sport field was established to label the images of the sport field according the annotation words in an image training set. Then, the traditional CMRM was adopted in the training images to complete the basic image annotations and obtain the image annotation result based on a probability model. Finally, the graph learning was used to refine the basic image annotations based on ontology concept similarity, and the top N keywords in the probability table for each image were chosen as the final annotation results. Experimental results show that the recall and precision of the proposed model are improved as compared with those of the traditional CMRMs.

Automatic Kayser-Fleischer(K-F) ring measuring systems are able to help the diagnosis of the Wilsons disease. As traditional diagnosis methods for the Wilsons disease ignore the influence of illumination and the corneal arcus on the detection precision, this paper designs an automatic K-F ring measuring system according to the (K-F) ring distribution in a color image. With the system, the Region of Interest(ROI) of an iris image was extracted by pretreatment. Then, the boundary of K-F ring was detected by the integral optimization of gradient algorithm and boundary tracking algorithm. To reduce the influence of the illumination on the validity of the algorithm, an illumination detection model was established to improve the robustness of the measuring system. Finally, the width features and color features were defined to remove the effects by normal images and corneal arcus. 2 234 iris images collected from our database were analyzed and the analyzed results indicate that the classification accuracy of the proposed K-F ring measuring system reaches 98.4%. It still has good performance if the patient has the corneal arcus.