The auto-focusing precision of microscopy has great influence on the performance of machine-vision-based precise measurement. A comprehensively quantitative evaluation method on image auto-focusing technique in a microscopic vision environment was researched. Several kinds of evaluation indexes were proposed, and the unbiasedness, unimodality, spatial resolution etc. of 13 groups sharpness functions were comprehensively evaluated in a microscopic vision condition. Then variance function and Brenner function were chosen to calculate the sharpness functions in coarse and fine focusing processes respectively. A modified Mountain Climbing Searching (MCS) algorithm was proposed to implement the micro-automatic focusing. As comparing to common MCS method, the modified method significantly improves the time consuming and increases the repeatability by about 24%. Finally, the developed auto-focusing algorithm was integrated into the system and was applied to the measurement of armature gap in a servo solenoid valve. The results show that the standard deviation of measurement is 1.9 μm, the precision is similar to that of the universal tool microscope, and the efficiency is significantly improved. Moreover, the system was also utilized for dynamic characteristic detection of gaps in the solenoid valve under the condition of power up, the relation between driving current and armature gap is obtained, which provides a reliable evidence for in-situ micro-assembly.

To comprehensively implement the UV atmospheric sounding of nadir and limb fields, the principle of simultaneously sounding of nadir and limb fields was analyzed and technical specifications were analyzed. Then a prototype of multiband UV atmospheric sounder with two fields was designed and developed. The optical system of the prototype is composed of a fore of lens, an annular lens, a relay of lens, and a narrowband filter. The central wavelengths of three working spectral ranges are 265 nm, 295 nm, and 360 nm, the bandwidth is less than 20 nm. The nadir field and the limb field are 10°, and 360°(141.8°—146.6°), respectively, the focal length is 5 mm, and the F number is 1∶3.3. Three wavebands were detected individually by switching UV narrowband filters. The total mass of the prototype is approximately 3 kg, and its volume is about Φ90 mm×300 mm. The spatial resolution and illuminance uniformity of image plane of the prototype were tested. The test result demonstrates that the static MTFs of nadir field and limb field are 0.25 and 0.22 respectively at a characteristic frequency of 38.5 lp/mm and the illuminance uniformity of image plane is 94%. Both MTFs and illuminance uniformity are better than that of design requirements. The prototype shows a smaller volume and a lighter weight, and satisfies the miniaturized and lightened requirements of spaceborne instruments.

To make effectively use of the polarized light information of the whole horizon and to explore the mechanism of bionic polarized light navigation, a polarization vision sensor was designed. The polarization vision sensor based on four cameras was introduced and its calibration method was given. Then, the optimal estimation of the polarization state was derived based on least square algorithm. The skylight polarization pattern was analyzed based on the first order Rayleigh scattering model, the estimation of sun direction vector was translated into an optimization problem of finding the minimum eigenvector, and the positioning and orientation algorithm was derived. Finally, the theoretical analysis was verified by a static experiment and a rotation experiment. The results show that the measured skylight polarization pattern is consistent with the Rayleigh scattering model and the solar vector can be extracted successfully from it. In the static experiment, the maximum error of the solar zenith angle is about 0.4 °with the standard deviation of 0.14 °and the positioning error is about 68.6 km based on the observation data within one hour. In rotation experiment, the maximum orientation error is about 0.5° with the standard deviation of 0.28 °. This study reveals the mechanism of bionic navigation based on polarized light, and provides a theoretical basis for its applications to the bionic polarized light navigation.

On the excellent optical performance of grapheme, a graphene-based all-fiber electro-absorption modulator was constructed. The structure of the modulator was designed, and simulation and experiments for the modulator were both conducted. Firstly, the relationship between the chemical potential and the permittivity was researched according to the graphenes optical properties, then the graphene-based all-fiber modulator was designed, and the dependence of the effective refractivity of the modulator on the chemical potential was analyzed. The proposed graphene-based all-fiber electro-absorption modulator was simulated and was compared to traditional semiconductor modulators. Finally, the graphene-based all-fiber electro-absorption modulator was developed and its performance was tested. The simulation results show that the modulation efficiency of the quadri-layer all-fiber graphene modulator reaches up to 0.233 dB/mm, and the footprint has been reduced to only 12.9 mm with the swing voltage lower than 0.2 V. The experiments demonstrate that the proposed modulator is feasibility. Limited by the practical conditions, the tested half-wave voltage is about 120 V and the 3 dB bandwidth is around 100 MHz. As compared with traditional graphene-based waveguide modulator, the proposed modulator shows a smaller half-wave voltage, higher modulation efficiency, and a smaller size. Moreover, it has a lower insert loss and is suitable for all-fiber systems in future.

A rapid volume measurement method for a bulk cargo was proposed to overcome the shortcomings of traditional bulk measuring systems in poor adaptability for yard environments, longer complete time, lower efficiency and complex operation. Then a volume measuring system consisting of a 2D laser scanner, a difference GPS and an attitude measuring system was designed. For this system, the laser scanner was used to measure dynamically the surface geometry information of a bulk, the attitude measuring system was taken to measure the space attitude data of the scanner in real time and the GPS was utilized to obtain the 3D position of the scanner in measuring processing. Finally, the data fusion was used to calculate 3D point cloud of the bulk cargo and to obtain the volume of the bulk cargo. On the basis of the feature of single scanned outline, an algorithm to extract the down edge of the bulk cargo effectively was proposed to remove the measurement error from ground point cloud and a projection triangulation volume algorithm was used to calculate the volume of bulks point cloud. Experiment shows that the system could complete the measurement of bulk cargo with a size of 69 m3 in 30 s, the average relative error is 0.42% and the repeated measurement error is 0.41%. In a reality bulk measurement, the measurement on about 31 500 m3 takes less than 10 min, and the average repeated measurement error for 4 coal bulks is 0.74%. It shows that the system could measure the volume of bulk cargo quickly and easily on maintaining accurate measurement.

For the profile deviation of a stage mirror itself and the profile change caused by the alignment of 2D stage position, an online detection method for the profile deviation of a nano precision 2-D stage mirror was presented. The basic principle of detecting profile differential data with two interferometers was proposed. The influences of zero error and integral error on the measurement were analyzed and a method to improve the measuring precision was put forward. Two sets of original data of the mirror profile were obtained by two sets of detectors with different spans composed by three interferometers. Through the relationship between the two sets of data, the detail profile deviation between the spans was modified. The theoretical calculation, simulation and experiments for the proposed method were performed, and obtained results were compared with that of off-line measurement with a Zygo interferometer. The results show that the most difference in test results is between ±10 nm, and the trend has better consistency, which verify that this measuring method could measure the profile deviation of stage mirrors correctly and could restore its surface errors truly.

Multiple Input Multiple Output (MIMO) systems can improve the channel capability of an Optical Wireless Communication(OWC) system without adding extra frequency resources and emitting powers, but the space correlation between the channels will impact the system performance. This paper researches the effects of space correlation on the Bit Error Rate (BER) of an optical MIMO system in log-normal fading channel. In view of the Pulse Position Modulation (PPM), the log-amplitude fading matrix in different correlation mechanisms and maximum likelihood detection rule were derived by using an exponential correlation model in the log-normal fading channel. On the basis of this, the upper bound of the BER of correlated OWC MIMO system was deduced by adopting Wilkinson method. Finally, the impact of space correlation on BER of the OWC MIMO system was analyzed further. The results show that space correlation deteriorates the performance of OWC MIMO system and the BER of the system will be worsen by the space correction with the increase of the number of the antenna. It suggests that antennas should be placed reasonably to reduce the space correction between the antennas, so to better play the advantages of MIMO systems.

To detect the shape of a Fiber Bragg Grating(FBG) sensor in a colon, a coordinate fitting method based on Frenet frame was proposed to improve the precision of the FBG curve reconstruction algorithm. Firstly, four FBGs perpendicular to each other were clinged to one Shape Memory Alloys(SMA) substrate. Hence, one FBG curve sensor with a diameter of 3 mm and a length of 900 mm was formed. Then, this sensor was put on a millimeter-square graph paper and a cylinder respectively to accomplish the reconstruction of 2D and 3D curves. Furthermore, every reconstructed value of data collection points was read from the curve reconstruction software and reconstruction errors were acquired by the contrast between the reconstructed values and the standard values read from the coordinate paper. Finally, the merit degree between two fitting methods was obtained. Experimental results show that the proposed coordinate fitting method based on Frenet frame effectively improves the reconstruction accuracy. When the FBG achieves their ultimate strain, the proposed fitting method can reduce the average error by 3.5%—5.5% at single direction, which lays a foundation for the precision upgrading of the FBG curve sensing system.

A high accuracy calibration method for the installation parameters between 3D Terrestrial Laser Scanner(TLS) and external-installed digital camera was proposed. Firstly, the intrinsic parameters were calibrated by existing commercialized close-range photogrammetry products. Then, by taking reflector target centers as input control points of 2D and 3D matching, the direct linear transformation algorithm was expanded in applications and the intrinsic parameters were taken as known coefficients to solve the extrinsic parameters. Finally, the extrinsic parameter of every camera shooting angle was calculated according to the relationships between different shooting angles to get the mapping relationship between point cloud and multi-angle images. The test experiments were performed, and the accuracy and feasibility of the calibration method were verified by obtaining the difference between point pixel coordinate and real measuring result with installation parameter calibration results. The experimental results indicate that the calibration accuracy reaches 1 pixel and the mapping relationship between point cloud and multi-angle images is correct, which satisfies the requirements of texture mapping and accuracy by applying images as auxiliary data in feature extraction.

For the surface measurement of optical elements, a surface measurement method by the randomly phase shifting interferometry of measured element was proposed to reduce the cost of phase shifting interferometer and to avoid phase-shifting error caused by an aged phase shifter. A micro-displacement driver was used to drive a measured element to move on a friction type air-bearing slider to implement the random phase shift, meanwhile, several interfere grams were collected by a camera. Then, the interferograms were processed by least-square iteration algorithm and the phase distribution of the measured element surface was iterated. Finally, the surface measurement result was calculated by a series of data processing and the surfaces of measured elements were obtained. To verify the feasibility of the proposed method, a Fazi phase shift interferometer was improved and a concave spherical mirror and a plane mirror were used as measured elements to perform the comparative experiment between the articles method with the traditional phase shifting method on the same instrument. Experimental results indicate that when the laser wavelength λ is 632.8 nm, the PV difference and the RMS difference between the two results are only 0.001λ and 0.002λ, respectively for the concave spherical mirror. Moreover, those between the two results are only 0.002λ, and 003λ, respectively for the plane mirror. These surface data are basically consistent. Experimental results show that the measurement method avoids the phase-shifting error caused by the aged phase shifter, and it has high accuracy and low cost.

To compensate the image motion of a space camera with large field, a general image motion velocity field model was established by combination of coordinate transformation and attitude dynamics for an off-axis three mirror anastigmat camera. In modelling, the effect of off-axis angles of three-mirror system on the image motion was taken into account, and the formulas of image motion velocity field in the off-axis three-mirror system were derived. By taking a space camera with large field for an example, the distribution characteristics of imaging motion velocity field under three typical imaging attitudes were analyzed. The influence of three axis attitude stability on imaging quality of the satellite was studied. The results indicate that the decline of satellite attitude stability, especially the pitching attitude stability, will lead to the decrease of dynamic Modulation Transfer Function (MTF)of focal plane. And the decrease is more obvious when the integration stages are increased. Moreover, the satellite attitude should be more stable when the camera is imaged in a side rolling with a large angle. For the space camera with 96 integration stages and the reduction of MTF no more than 5%, the satellite three-axis attitude stability should be controlled under 0.001(°)/s. the experiment results confirm the above analysis of satellite stability, verify the accuracy of the proposed image motion velocity field model and provide a reliable basis for the image motion compensation of space cameras with large fields.

This paper focuses on the reliability evaluation of photovoltaic(PV) module products under accelerated degradation conditions. A method to estimate the pseudo failure life distribution of PV modules based on the Generalized Lambda Distribution(GLD) was proposed. Firstly, the determinative coefficient test method( R2) was used to select the best accelerated degradation path to obtain pseudo failure life values of samples. Then, the bootstrap method was taken to generate bootstrap samples to expand the sample group and to build a pseudo failure life distribution model based on GLD, which could truly reflect the pseudo failure life distribution of PV modules under different acceleration conditions without presetting the prior distribution, Finally, by taking accelerated performance degradation data of a 18 W small power Mono-Si silicon PV module as an example, the pseudo failure life distribution and reliability of PV modules were estimated under accelerated degradation conditions. The results show that this method obtains pseudo failure life distribution in accelerated degradation conditions, and reliability curves and the experimental data results fit well. Moreover, the absolute error of fitting curve is basically in 300 h, and the relative error is not more than 15%. These results meet the demand of engineering forecast precision.

The surface characteristics of reaction-bonded SiC (RB-SiC) ceramic with Electrical Discharge Diamond Grinding (EDDG) were explored. Three kinds of processes of EDDG, Electrical Discharge Grinding (EDG) and Conventional Grinding (CG) were taken into the experiments of RB-SiC grinding. Then the surface roughnesses of the RB-SiC were measured with a laser scanning confocal microscope, and their surface morphologies and micro-cracks were observed with a scan electron microscopy. The grinding characteristics of the RB-SiC with EDDG were analyzed and compared to that with other two methods. The results show that the surface roughness value of RB-SiC is 0.214 9 μm with the EDDG, which is better than that with EDG but slightly poorer than 0.1956 μm with the CG. For the surface morphologies of the RB-SiC, it shows obvious discontinuous scratches with the CG, and is mainly composed of discharge craters with the EDG machining. However, both scratches and discharge craters are existed on the machined surface with the EDDG. Moreover, the grinding cracks and grain boundary cracks are found on CG surface and only hot cracks exist on the EDG surface. But the micro-cracks on EDDG surface are divided into grinding cracks and hot cracks, and the latter can be ground with diamond grits. With the investigation of comparative experiments, the results show that different machined surface characteristics of RB-SiC ceramic with EDDG process from that with EDG and CG processes.

As the micro-vibration of a satellite platform restricts the imaging quality of a high-resolution space optical remote sensor, this paper designs a six DOF(Degree of Freedom) platform for the ground test of optical remote sensor alleviation margin in satellite micro-vibration environment. The kinematics and dynamics models of the platform were constructed, and the transfer function, Simulink model of a voice coil actuators were derived. Based on the models, the platform with six DOFs was manufactured. A confirmatory experiment on the vibration acceleration control accuracy of the platform was carried out, in which the micro-vibration frequency of the typical satellite was taken as the input signal. The results show that the relative error of output acceleration has been controlled in 7% in frequencies from 7 Hz to 40 Hz. The platform takes the parallel construct of the stewart model, it has advantages in simpler structure, bigger stiffness and a controllable vibration source, and obtained results meet the requirements of the ground test applications.

A novel asymmetrical inertial piezoelectric rotary actuator with a bias unit was proposed for improving the output performance of inertial piezoelectric actuators. On the asymmetrical clamp, the bias unit was designed. To research the effect of the bias unit on the output performance of piezoelectric rotary actuator, a mechanical model equation of the mechanism was initially established, and the dynamic characteristics of the actuator were also derived and simulated. Then, a prototype was designed and fabricated, corresponding experimental system was built to conduct experiments and experimental results were compared with that of the actuator without the bias unit. The results show that the output angular velocity peak of the actuator is at the offset distance of 15 mm. Compared with the actuator without the bias unit under the driving voltage of 100 V, 23 Hz, the actuator with the offset distance of 15 mm improves the maximum angular velocity from 3.48 rad/s to 5.39 rad/s, enhancing by 54.88% and the highest driving moment from 2.41 N·mm to 3.62 N·mm, enhancing by 50.2%. In addition, its heavy bearing capacity at stable motion attains 1 300 g under the driving voltage of 100 V, 4 Hz. Both theoretical and experimental results indicate that the designed actuator hold advantages in larger angular velocities and inertial driving moments in comparison to the one without the bias unit.

To effectively compensate the hysteresis nonlinearity of piezoelectric ceramics, a modified PI model based on STOP operator was proposed to avoid the complex processing in solving inverse model and time consuming in interpolating method of the traditional PI model based on PLAY operator. Firstly, traditional PI models based on PLAY operator or STOP operator were introduced. Then, modified PI model based on STOP operator was established by taking an expecting displacement as the input and a control voltage as the output, and the model was used as a feedback controller to compensate the hysteresis effect of piezoelectric ceramics. To balance the ability of local optimization and global optimization, the particle swarm optimization algorithm was improved to identify the weights of operators with different thresholds. Finally, the modified PI model was used to verify experimentally the compensating effects for the hysteresis nonlinearity. Two groups of experiments were carried out, and the results show that the hysteresis has been compensated well by modified PI model with the error no more than 1% no matter the input is continuous or random. It concludes that, the modified PI model based on STOP operator is of great value in the field of piezoelectric ceramic control.

The precision deposition of micro/nano patterns printed by Electrohydrodynamic Direct Writing (EDW) technology was explored and the EDW for orderly nanofibers was implemented by the straight stable jet between a spinneret and a collector. The deposition behaviors of EDW jet on stationary and moving substrates were investigated. The effects of process parameters on the position errors of EDW patters were also studied. The experiments show that the jet is bended into a spiral structure by the inner stress and charge repulse force, and then the nanofiber is guided to form a three-dimensional fibrous microstructure on the stationary substrate. By increasing the velocity of collector, the bending process of charged jet can be overcome and the straight line nanofiber without spiral coil is direct-written on the substrate. The multi rectangle wave and square wave nanofibrous patterns are direct-written according to the designed pattern, respectively, and the dimension error between the direct-written nanofibrous pattern and designed pattern is also analyzed. The results indicates that the position error of direct-written nanofibrous pattern increases with increasing the velocity of collector, the distance between spinneret and collector, applied voltage, and the motion distance of collector. Moreover, by optimizing the experimental conditions and design parameters, the position error of direct-written fibrous pattern can be less than 10 μm. It concludes that the precision deposition of micro/nano pattern is benefit to promoting the control level of EDW technology.

Fast Steer Mirrors(FSMs) should possess different effective bandwidths and close bandwidths under different situations. So this paper establishes a system model based on a piezoelectric FSM control system, and optimizes the control strategy for the FSM system by analyzing the jitter of an optical axis. Firstly, close loop Bode figure of the system was measured, the simulated annealing algorithm was used to achieve the modulation transfer function of the system model. By combining simulated annealing algorithm and the identified model, an optimal PID controller to meet all kinds of applications was presented. Finally, the step response tests were used to verify the correctness of the identification model, and the closed-loop tests were taken to validate the effectiveness of the optimal controller. Experimental results indicate that the indentified model fits the FSM system very well in low-middle frequency stage, its step response is similar. The system controlled by the optimal controller has a closed-loop bandwidth of 70 Hz, an effective bandwidth of 35 Hz, and the tracking accuracy is improved by 47%, meeting the requirements of FSMs in the current environment. It concludes that the system has good low frequency tracking ability and high frequency interference rejection ability, showing high tracking precision and low device loss.

For identification of the coupling stiffness of MEMS (Micro-electro-mechanical System) gyroscopes, a identification method was proposed based on the frequency response characteristics of the drive-axis, sense axis, drive-to-rotation coupling and rotation-to-sense coupling. A dual-mass linear vibrating MEMS gyroscope with decoupled drive-to-sense and sense-to-drive displacement was designed. Based on simplified stiffness characteristics of the beams, the dynamic planar movement equations of the gyroscope were established and the drive-axis, sense-axis, drive-to-rotation and drive-to-sense transfer functions were derived. According to the coupling model, the sources of stiffness coupling were attributed to the stiffness error of specific beams. The drive-to-rotation coupling stiffness could be identified by the ratio of drive-to-rotation coupling to drive-axis frequency responses, and rotation-to-sense coupling stiffness could be identified by the ratio of drive-to-sense to sense-axis frequency responses. The frequency responses of the gyroscope were investigated by the proposed coupling stiffness identification method, and results show that coupling stiffness coefficients by drive-to-rotation and rotation-to-sense for the tested gyroscope are 0.14 N and 0.054 33 N, respectively. It concludes that the identification results provide references for laser trimming of the beams for gyroscopes.

A novel inverse feedforward control algorithm was developed based on inverse Bouc-Wen feed-forward control and feedback control for improving the trajectory tracking performance of a Piezoelectric Actuator (PEA). A Bouc-Wen hysteresis dynamic modeling for the PEA was established, and system parameters of the Bouc-Wen model were identified by Particle Swarm Optimization (PSO) method. Then, a feed-forward compensation control method was proposed based on hysteresis Bouc-Wen model. Finally, the inverse feed-forward control method combining the PI feedback control with feed-forward control were proposed to control the piezoelectric actuator. An experimental platform was developed based on dSPACE system. The hysteresis experiment results show that the hysteresis error and relative linearity of the proposed method is almost zero and 96.5%, respectively. The trajectory tracking experimental results show that the maximum tracking error and RMS tracking error of the proposed method are 0.180 5 μm and 0.055 4 μm, respectively, obtaining the high tracking performance by 10－8 m. As compared with open loop control, feedforward control, PI feedback control, the proposed inverse feedforward control algorithm compensates basically hysteresis nonlinearity of the PEAs and shows good trajectory tracking performance.

To improve the precision of rapid positioning of a piezoelectric positioning stage, a positioning model is established to characterize the relationship between driving voltage and output displacement of the platform. As the piezoelectric positioning stage for precision position in the fast and large stroke is not only affected by the hysteresis characteristics of the piezoelectric ceramics, but by the dynamic characteristics, this paper uses Bouc-Wen model to describe the hysteresis characteristics of piezoelectric ceramics, combines with the dynamic characteristics of the piezoelectric positioning stage to establish model, and lets the model reflect the dynamic characteristics and hysteresis characteristics of the platform. To verify the correctness of the model, the experimental equipment based on the piezoelectric micro displacement platform and the related driver is built, and model is verified by experiments, and a secondary development of monitor-control program is performed.The results show that, compared with the previous Bouc-Wen model,when proposed model in the maximum displacement output is 40 μm and the input voltage frequency is 40 Hz,the maximum error is reduced from 3.04 μm to 0.67 μm and the maximum relative error is 1.68%.The results show that the proposed model can better simulate the hysteresis characteristics and dynamic characteristics of the piezoelectric stage, and greatly improve the accuracy of the piezoelectric micro displacement platform in the fast and large stroke positioning.

To optimize the deploying performance of folding wing mechanism for an aircraft, this paper analyzes the folding wing mechanism and proposes a method to optimize the deploying performance of the mechanism. A theoretical model and a dynamic simulation model for the deployment of folding wing mechanism were established, the deploying processing of folding wing mechanism was analyzed and the folding time of the mechanism was given by using the theoretical model. Then, the effect factor on the deploying performance of the mechanism was analyzed, and the orthogonal trial method was used to optimize those structure parameters to obtain a optimized designed scheme. Finally, the mechanism was simulated and optimized by the optimal scheme and the deploying performance of the mechanism was measured by experiments. Experimental results show that the folding time of the mechanism is 0.128 s and the structure stresses at the measuring points are 92 MPa and 80 MPa respectively. Moreover the dynamic model test results of the mechanism show that the folding time of the mechanism is 0.12 s and the structure stresses at the measuring points are 85 MPa and 96 MPa, respectively. The difference of the experimental and simulation results is within 5%. It indicates that the optimal folding wing mechanism promotes the flight performance of the aircraft while the mechanism also basically satisfies the design requirements like the stability, reliability, fast speed, small stress and the impact force.

On the basis of the measuring principle of time grating sensors, a new variable coupling time grating displacement sensor was proposed by employing a high-frequency pulse as measurement basis to improve the precision of displacement measurement. The precision displacement measurement of the sensor was implemented by changing the coupling state of excitation and induction coils to output a traveling wave signal for inducing the displacement change. By modeling and simulations, the displacement error characteristics of the sensor when sensor-head was in different attitudes were obtained. The harmonics analysis was performed, and the influence of different sensor-head attitudes on different resonances of the displacement measurement errors was obtained. A sensor prototype was designed based on the model and then experiments were carried out. Simulations and actual experiments indicate that different sensor-head attitudes mainly influence 1st, 2nd and 4th harmonics, and other harmonics are more sensitive to pitch attitudes than other attitudes of the sensor-head. When the sensor-head is in a good attitude and the gap between ruler and sensor-head is 0.3 mm, both experimental analysis and simulation result show that the measurement error is about ±18 μm. Simulations and actual verification experiments are in agreement well.

A detail preserving multi-exposure image fusion algorithm was proposed to address the problem of the loss of visual details and ghost artifacts in traditional multi-exposure images. Firstly, three image features, image details, exposure brightness and color information, were calculated. In which, the image details were obtained by using a guided filter, the each exposure intensity was weighted by a Gaussian function and the color information was measured by color saturation. Then, the difference maps and correlation coefficients were used to detect the motion objects in dynamic scenes and the focused weight map of static and dynamic scenes were calculated respectively by using feature indexes and detection results. In order to remove the noise effect, a recursive filter was used to correct the focused weight image, and the focused image was obtained by a pixel-by-pixel weighted sum of the input images. Ten kinds of multi-exposure image sequences were tested in the experiments and obtained results were compared with that of six kinds of traditional methods. The experimental results demonstrate that the proposed algorithm exhibits good visual appearance and preserves more details. It also effectively removes ghost artifacts in dynamic scenes. It concludes that the proposed algorithm is better than 6 classical methods and it produces desirable images in both static scenes and dynamic scenes.

Since Kernelized Correlation Filters (KCF)tracking algorithm is sensitive to feature selecting and unable to estimate object scale, this paper researches the KCF tracking algorithm based on feature extraction and scale adapting. A scale adaptive KCF tracker by using HHS-OG (Histogram of Hue Saturation and Oriented Gradient, HHS-OG) feature was proposed to improve the tracking performance of the KCF tracker. Firstly, the HSI color space was studied. By taking the complementary of color and gradient in an image, a novel HHS-OG feature focused color and gradient features was proposed to improve the discrimination ability of the KCF algorithm to backgrounds and targets. As the KCF algorithm is unable to process the changed scale, a set of scale factors were used to sample image patches in the detection stage of tracking and the generated corresponding filter response maps were used to estimate the optimal target position and scale. The proposed tracker was tested on a large tracking benchmarks with 50 video sequences. Experimental results show that the tracker runs at a high speed of 37.5 frame per second and has a significantly improvement of 5.4% in representative precision score and 10.1% representative success score. The HHS-OG feature has good discrimination ability for backgrounds and targets and has robustness for target tracking. The scale adaptive strategy is effective for improving tracking performance.

Some of the traditional single-frame super-resolution (SR) reconstruction algorithms can not get good reconstruction results, although they learns many different types of images. Therefore, a super-resolution method combined with the Support Vector Regression (SVR) and raster-scan actions was proposed. Firstly, image patches were extracted from a group of high resolution (HR) images and the corresponding low resolution (LR) edition by the raster-scan actions, and input vectors and pixel vectors were taken out from the patches. The Log algorithm was used to determine that those patches were belong to high-frequency space or low frequency space then to construct the high and low frequency vector pairs. Then, those optimized vector pairs were trained by the SVR and two dictionaries in high/low frequency spaces were built eventually. Furthermore, input vectors were extracted from tested LR images in high/low frequency space, and the SVR tool was used to predict the SR pixel labels and the predicted pixels were added to bicubic interpolation image based on LR edition. Finally, the SR image was obtained by post-processing the previous image. In comparison with other algorithms, experimental results indicate that the proposed method provides good visual effects. It enhances its Peak Signal-to-Noise Ration (PSNR) and Structural Similarity Index Measurement (SSIM) by 3.1%-5.3% and 1.5%-8.1% on different images, respectively as compared with bicubic interpolation method.

An improved Hybrid Spill-tree algorithm based on the signed method defined as Signed Hybrid Spill-tree (SHSPT) was proposed for target matching of remote sensing images. For establishing data and preprocessing, a data separate method based on a center point was proposed, the separated data were extracted by defining the center of dense data, and the edge data were abandoned. In the feature matching, binary array were used to express the data space and to mark the feature vector. Then, the bit operation was used to compute the distance between the feature vectors and to shorten the computing time. Finally, the feature matching algorithm was improved. The average value of the feature distance was used to replace the secondary characteristic distance from the Scale Invariant Feature Transform(SIFT)matching algorithm to obtain more matching points. The test results show that the computer memory by proposed algorithm is reduced 68% than that of traditional hybrid spill-tree method, and matching accuracy is closed to that of the traditional one. In addition, the method reduces 32.8% matching time. It solves the problems of remote sensing images in larger data amounts, higher dimensions, longer matching time and larger computer memory.

For the phase retrieval in fringe projection 3D shape measurements, a new fringe projection phase retrieval method based on variational mode decomposition was proposed. Firstly, the projection fringe pattern was decomposed into a background part, a fringe part and a noise part by the development of variational mode decomposition model and the minimization of the model. Then, the fringe part was processed by Hilbert and arc tangent transform to obtain a wrapping phase, and by quality guided phase unwrapping and Zernike carrier removal to acquire the final absolute phase. Simulation and experimental results show that the phase error by the proposed method is 3.14×10-4, smaller than the errors 3.30×10-4 and 6.52×10-4 that respectively obtained by Fourier transform method and continuous wavelet transform method. The proposed method is superior to the Fourier transform method and continuous wavelet transform method in the process of projection fringes with edge information, providing more accurate results, and is more effective for the application of the three dimensional measurement of objects with discontinuous and abrupt changes.

To compensate the axis errors between rotation axis and vertical axis when a digital zenith camera was used in celestial positioning, a incline angle compensation method for the digital zenith camera was proposed. On the principle of incline angle compensation, the scale coefficients and the cross angle of two axes were taken into account to modify the inclinometer values. Then, a calibration method was proposed for the parameters of the two-axis inclinometer. The influence of rotation angle on calibration parameters was analyzed, and some experimental data were used to verify the calibration method. The result demonstrates that the rotation angle directly affects the calibration of CCD installation angle. Moreover, the introduction of inclinometer parameters improves the positioning accuracy. If the error of rotation angle is below 2°, the influence caused by inclinometer parameters can be ignored.

A threshold decision directed sparse channel estimation (TDDSCE) method was proposed for a Single Carrier Block Transmission(SCBT) system, which can take advantage of the sparsity inherent in high rate wireless communication channels, and improve the channel estimation accuracy. Firstly, a pilot sequence was used to perform a least square (LS) channel estimation, then a decision threshold was derived by obtained channel estimated values. The channel taps whose values were smaller than the decision threshold were forced to be zeros, only the channel taps whose values were above the decision threshold were reserved, so that the sparse channel estimation accuracy was greatly improved. A simulation experiment was carried out based on a COST 207 rural area channel profile, and the results show that the performance of proposed method is most close to that of the known channel case, and its bit error rate (BER) reaches 5×10-4 when the signal to noise ratio is 20 dB. However, the traditional LS channel estimation method only achieves the BER of 3×10-2. The method improves the sparse channel estimation accuracy and reduces the complexity of the SCBT system, and its feasibility is verified by obtained results.

To solve the invalid tracking of a human target caused by appearance variations due to large angle change of the target in a robot mobile tracking, a multi-template regression weighted mean-shift algorithm was proposed. The algorithm could implement the target tracking by building a multi-template model of the target and applying mean shift. Firstly, the template set was obtained according to the results from mean shift procedure of the last frame and the coarse location information of head-shoulder model of a current frame, by which the position and angle variation of the target person were included. Then, the multi-template regression weighted mean-shift algorithm was used to determine the precise location of the target person. The regression model was introduced to multi-template mean shift to implement a map from color-texture feature to the similarity of target model to limit the number of templates and to ensure the real-time performance of the target detection. Finally, the proposed algorithm was verified by videos and robot tracking tests. The results show that the image average treatment time is 86.4 ms/frame, which satisfies the requirement of person tracking for a mobile robot. The method solves the appearance variation problem of targets in tracking processing and improves the robustness of human targets to its feature variations.