In order to accommodate as many subapertures as possible in the high-speed wavefront slope calculation when hardware resources were limited, an accumulator-based wavefront slope processor was proposed. The computational core of the processor is an array of subaperture spot centroid calculating element.According to the decomposition of 2D moment calculation and the recursive procedure of 1D moment calculation,it can substitute the multiplications between the pixel grayvalues and its coordinates with several sum operations to obtain the low order 2D geometric moments required in centroiding. The calculating element simply consists of five accumulators, and the cost has been decreased because no multipliers are needed. Experimental results indicate that when it is implemented in a Field Programmable Gate Array(FPGA) at a clock frequency of 100 MHz, the proposed architecture can obtain gradients of all subapertures in 22×22 Shack-Hartmann with the latency no more than 0.33 μs and the error less than 0.002 pixel. The new design has reduced the hardware resource by 40% as compared to that of the multiplier-based architecture. Furthermore, the original multiplier-based processor can be updated for Shack-Hartmann sensors and can obtain the subapertures twice as much as that of original one without additional hardware resources. It realizes the high speed measurement of wavefront with a high accuracy.

To resolve the contradictions between wide Field of View(FOV),large exit pupil and compact，lightweight,a hybrid refractive/diffractive optical system using Free Form Surfaces(FFSs) for a Head Mount Display(HMD) was designed.The dispersion characteristics of diffractive elements were used to correct the chromatic aberration,and the appropriate types of FFS were selected to correct the off-axial aberrations(astigmatism，coma and distortion). Furthermore,the FFS prism was made of plastic materials to form an optical structure without rotational symmetry to achieve compact and easy to align functions.Obtained optical system can offer the eye relief of 17.4 mm,exit pupil of 4 mm and the FOV of 55°.Moreover,the RMS diameters of spot diagram are less than 30 μm,the MTF are higher than 0.1 at the spatial frequency of 30 lp/mm across the entire visual field,and the maximum distortion is less than 10%. Experimental results indicate that the optical system has perfect performance and can meet the demands of eyepieces for aberration characteristics.

The laser beam drift is one of the key factors that influence on the accuracy of laser measurement of straightness errors, and a novel method for straightness error measurement based on common-path compensation for the laser beam drift was proposed. The measurement principle and system configuration were given, influences of the laser beam drifts caused by different factors on straightness error measurement were analyzed，and a compensation model was established.After compensation,the results show that the variation of emission beam is reduced from 28.4 μm to 5.6 μm and from 21.6 μm to 5 μm in X direction and in Y direction, respectively,and the variation of beam drift due to temperature gradient is from 65.7 μm to 8.9 μm. Both experimental results and theoretical analysis show that the common-path compensation for the laser beam drift method can decrease greatly the influences of the laser beam drift on the straightness error measurement, and can enhance its measurement accuracy.

Circular Subaperture Stitching Interferometry (CSSI) and Annular Subaperture Stitching Interferometry(ASSI) were expounded to realize the test of large aspheric surfaces without the null optics. The theories and principle of the technique were analysised,and the synthetical optimization stitching model and an effective stitching algorithm were established based on homogeneous coordinate transformation and simultaneous least-squares fitting. The softwares of CSSI and ASSI were devised, respectively, then the experiment was carried out for a Φ350 mm hyperboloid by CSSI, ASSI and a null test, respectively.Experiments show that the differences of PV and RMS errors between null test and CSSI are 0.031λ and 0.004λ, and those between null test and ASSI are 0.028λ and 0.006λ, respectively,while their synthesized surface maps are consistent to the entire surface map from the null test. The technique proposed in this paper provides another quantitive measurement for testing large aperture aspheres besides the null-compensation method.

For the purpose of providing a method to estimate the effect of vibration on the optical performance of a burning diagnosis optical system,a numerical calculation method for the misaligned optical system under interference vibration was proposed. Based on the finite element analysis software ANSYS, the physical model of misaligned optical system under interference vibration was developed and the transient dynamic analysis was inspired by a reflector displacement in the vibration response. Then,the dynamic response of the reflector was obtained by transient dynamic analysis. On the basis of the matrix optics, geometrical optics theory and the mechanic vibration theory, the vibration misalignment matrix of planar reflector was analyzed, and the vibration misalignment model was also built. Meanwhile, the method of simulating calculation of misaligned optical system under interference vibration was inferred. Furthermore,an experiment was carried out to validate the method. The results show the the relative error of simulating calculation and experimental data can reach 4.1% and 0.8% in the horizontal and vertical directions,respectively.

In order to correct the laser pointing errors caused by the axial errors for a level mounting laser transmitter system, a new laser pointing error correction model was built by using the correction method used in the pointing error correction of a photoelectric theodolite,and the error transfer behavior between shafts and optical path was investigated intensively. Firstly, the opto-mechanical structure and modeling theory were introduced, and the transformation matrixes of mirrors were presented. Then, by constructing two 3D coordinates in which the laser beam was considered as a unit vector, the laser pointing model was established by linear coordinate transformation and linear vector transformation. Furthermore,the undetermined coefficients in the laser pointing model were obtained by least square fitting.Finally, the TV tracking system whose optical axis was actually considered as an ideal one in the laser transmitter system was used to measure the laser pointing error. Experimental results indicate that the laser pointing precision of the level mounting laser transmitter system is up to 3.1″ and 9.7″ in 2 specific orbits and a zenith region, respectively, which satisfies the requirement of the system design for the accuracy.

With the aim to serve in bad environments and to meet the requirements of real-time and long-term service, a new kind of double parameters and high temperature-pressure fiber Bragg grating(FBG) sensor with a temperature-compensator was put forward. Firstly,the constant elasticity alloy was selected based on the special environments for the sensor. Then, a new kind of structure combined with a circular cylinder and a circular diaphragm was designed by taking the optimized constant elasticity alloy as the substrates. Finally, three experiments of temperature, pressure and temperature-pressure were performed. Experimental results indicate that the sensor not only has great temperature and pressure detection ranges, but also can realize the linear measurement within the ranges of 0~60 MPa and 0~350 ℃.The experiments offer the pressure sensitivities in 0.013 6 nm/MPa, temperature sensitivity in 0.020 1 nm/℃, and the static errors in 0.046% and 0.029%,respectively. These indicators can meet the requirements of practical engineering applications.

To reduce the influence of stray lights on the spectral measurement accuracy of a High Resolution Imaging Spectrometer(HRIS), sources and hazards of stray lights were analyzed. Based on the Point Spread Function (PSF) theory which is usually used to describe the stray lights of HRISs,the influence matrix and the correction matrix of stray lights were discussed. The measurements and specific correction programs were given, and the accuracies of measurement and correction were analysed. The results show that the method based on PSF can complete the measurement and correction of stray lights for the HRIS. For a pixel center wavelength incident light, the effective signal is only reduced no more than 1%, while the stray light signal is reduced at least by 99% as compared with the raw data. For both pixel center wavelength and non-pixel center wavelength incident lights, the corrected signals are reduced by about 7% at the pixel center wavelength and by nearly 25% at the non-pixel center wavelength, while the stray light signal is reduced by almost 50%. Finally, the theoretical reasons why the incident light wavelength affected the stray light correction were explained.

On the basis of Y loop elements, a new method to set circular slots on the periodic element to improve the narrow bandwidth of a Frequency Selective Surface (FSS) structure was presented. The FSS structure was calculated by Galerkin’s method in the spectral domain when the optimum radius was set to be 0.5 mm and the number of the circular slots to be 12, and the transmittance in the resonance frequency of 10 GHz has increased by 0.11 dB. The plane sample was fabricated by depositing coatings and a lithograph technology,and obtained measured values are in good agreement with the calculated values. The results show that the transmittance of this structure is about -0.62 dB at the resonance frequency,which has increased by 0.21 dB under the perpendicularity incidence compared with that of the Y loop without circular slot. Furthermore,the transmittances at TE 30°and 45°incidence are -0.66 dB and -0.81 dB, respectively,which has increased about 0.32dB and 0.27dB compared with those from the Y loop without circular slots.The bandwidth difference between the two structures is negligible, which are about 1 GHz and 0.8 GHz under TE 30°and 45°incidence, respectively. These results indicate that this method is an effective way to improve the transmittance of large periodic FSS structure at the resonance frequency.

To measure the rounded corner and diameter of a micro-drill, the fitting of the elliptical equation for margin projection of the micro-drill was researched.Firstly,the margin projective image was collected with a high precise automatic test system of PCB micro-drill. While the elliptical equation of margin projection of micro-drill was fitted, the sum of square distance from the sample points’ vector coordinates to the hyperplane determined by the elliptical coefficients was taken as an objective function. Then, a Sanger neural network with lateral connection was designed, in which the input signal were vector coordinates composed of sample points.Finally,on the basis of an adaptive minor component extracting method,the eigenvector of minimum eigen values was taken as the fitted coefficients of the hyperplane, i.e., fitted elliptical coefficients, and the rounded corner and diameter of the micro-drill were obtained. The proposed approach can meet the precision requirements of micro-drill testing, and is a novel application of adaptive Sanger network to measuring shape errors.

An improved method to fit the side wobble trace of a positioning setup to compensate the software was proposed to reduce the error from the mechanical swing of the focusing setup during focus plane adjustment.The alignment mark was positioned through the sub-pixel correlation template matching algorithm and the validity of the algorithm was investigated, which shows that the resolution of the algorithm is better than 0.1 μm. The positioning algorithm mentioned above was used to calculate the coordinates of the test mark while it moved up and down along with the focusing positioning setup and then the waving trace was characterized. Experimental results indicate that the mechanical swing trace can offer a good reproducibility, if the ascending and descending of the focusing setup are considered,respectively. Based on the experimental results, the model for the swing trace fitting was established and an error correction method was presented. Finally, experiments were performed to test the calibration accuracy of the model.Obtained results demonstrate that the error from the mechanical instability has been reduced to 1.29 μm from 2.84 μm, which can meet the overall alignment requirement of 2 μm.

A CMOS full differential interface circuit with low noise and high linearity was presented for closed-loop capacitive micro-accelerometers. Based on switched-capacitor detection, the circuit was designed to improve its linearity by a 0.5 μm n-well CMOS process technology. The simulation result shows that the proposed two-path feedback structure provides a good system linearity of 0.01%. The optimized designed PID controller was added into the system,which decreases the stabilization error effectively, increases the system responding speed by 31%, and the linearity by 66.7%. With a ±5 V supply and a sampling frequency of 64 kHz, the circuit can offer the equivalent input noise in 8 μg·Hz-12, system sensitivity in 1.22 V/g, system linearity in 0.03%, and the work range in ±2 g. These results prove that this circuit is suitable for applications of high performance micro-accelerometers to seism detection, oil exploration,etc.

A type of focal plane-curtain shutter was designed for a small focal plane aerial camera whose CCD image was orthogonal and vetical distribution due to the prism splicing. By taking the cog belt as the shutter curtains and the appropriate strip-shaped slit on the curtain as exposure slit, the exposure time of the shutter was determined by the velocity of the shutter curtains. The simultaneous exposure of two pieces of CCDs was implemented by the orthogonal layout and synchrodriving of two shutter curtains. According to the feedback of an encoder, the velocity of shutter motor was controlled by a DSP.Therefore, the shutter curtain ran at a steady speed and offered the exposure uniformity of CCD. Experiment results demonstrate that the shutter has a higher movement stability and a long life span in the range of continuous stepless adjustment exposure time (1/300-1/1 000 s), and its exposure error is less than 10%. Furthermore,by changing the shutter curtain and the slit width, the exposure time range can be extended. Obtained result shows that the performance of this shutter fully satisfies the requirements of area CCD aerial cameras.

The transient S-parameters of a capacitive shunt and electrostatically actuated millimetre-wave switch were researched.The transient change of these parameters during switching in the gap between switch beam and dielectric layer was derived by a one-dimension mechanical dynamic model from a published paper. Then, the transient change in the gap was used for simulating the transient S-parameters during switching in HFSS software. Finally,this method was applied to a specific switch. Obtained results show that the inserting loss S21 decreases slowly until the end of the pull-in process (from steady value -0.20 dB in Up-state to -1.02 dB over 9.11 μs, 97% of the pull-in time), and it increases quickly to nearly the ultimate value at the beginning of the release process (from steady value -20.1 dB in down-state to -1.16 dB over 1.09 μs, 5.5% of the release time).The transition time from a passing RF signal to a blocking RF signal is about the time of pull-in, and the transition time from a blocking RF signal to a passing RF signal is one order of magnitude less than the releasing time.

A combination measurement system for pitching angles was put forward based on a collimated laser, a single-axis inclinator and a CCD camera to improve the precision and stability of measuring pitching angles in engineering. The combination measurement model was introduced and its combination measurement principle was studied in detail on the basis of the coordinate rotation algorithm and the iteration method. The global sensitive analysis of each factor affecting the combination measurement was carried out by using Monte Carlo method.Obtained results show that compared with the direct measurement method with an inclinator widely used in pitching angle measuring at present, the proposed method can obtain much higher measurement precision and stability. A measurement experiment on a highly accurate three-axis turntable was performed,which indicates that the maximum absolute error is 0.3 mrad and the standard deviation is 0.16 mrad.These data are better than that from the direct measurement method with the inclinator,and meet the requirements of the engineering applications.

In order to meet the requirements of a tracking gimbal for the lower cost,shorter manufacture periods and higher interchangeable property,a new way of shaft transformation by using standard-bearings to replace thrust bearings was proposed. The axis error sources of using standard-bearings were explained, then according to the principles and characters of the tracking gimbal, an error eliminating method to make the axis sloshing become parallel moving was given. Finally,an experiment model of horizontal axis was designed, and the error of axis adopting zero level deep groove ball bearings was tested and analyzed by the Fourier harmonic analysis method. The experimental results indicate that the maximum error of shafting sloshing has been reduced from 6.126″ to 2.430″, which proves the the feasibility of the proposed error eliminating method. It concludes that the higher level bearings and right fixing plans can meet the precision requirements of small-scale tracking gimbals.

By utilizing the numerical solution and Finite Element Analysis (FEA) approach, the effect of the air gap beneath a heating membrane on the performemce of a micro heater was calculated and simulated. The thermal convection coefficient was acquired from a heating experiment. Then, a 1D Fourier heat transfer equation was derived.By using the Biot number calculated and the lumped-capacity solution, the model was simplified into a multi-layer thin slab one.Furthermore,the transient temperature response and stable thermal distribution of the air gap in thickness of 0 (pure Si substrate), 100, 200, 300, 400 μm and completely through (heating membrane) were compared under the conditions of heat convection and heat transfer. Calculation results show the climax temperature has increased approximately 17% by utilizing the heating membrane structure. The results of steady state and transient thermal-electrical coupled field FEA reveal that 200 μm air gap structure indeed enhances the climax temperature to 390 K and reduces the power consumption to 134 mW, which is coherent with the numerical calculation results and experiences.

A planar micro-spring based on a planar rectangular helical spring was designed for a low-gn micro inertial switch with the micro-spring in stiffness up to 0.1-1 N/m magnitude. The elastic coefficient formula of the micro-spring was derived with the Castigliano’s theory and the linear elasticity theory based on the assumption that the material property was uniform, homogeneous and isotropic, and then it was modified by the ANSYS finite element method. The micro-spring was fabricated by using the MEMS bulk micromachining process based on the multi-step ICP etching method,and the bulk of chip was about 7 mm×7 mm×0.3 mm in size. The theoretic analysis shows that the results agree well with the ANSYS simulation results, which demonstrates that the optimized formula can be directly applied to the optimization design of the micro-spring to simplify the design process. The nanoindentation test results show that the elastic coefficient of the fabricated micro-spring is about 0.554 N/m. The planar micro-spring has the advantages of small size, simple structure and easy fab-rication, and it can lay the foundation for the engineering applications of low-gn micro inertial switchs.

A test system was established to test the photoelectric characteristic parameters of thin film and module solar cells rapidly and accurately, and the control mode and test accuracy of the test system were studied. By taking the DSP chip TMS320LF2407 as the main chip, the modular and bus structures as the design idea, the obtained system show a faster response and a good scalability. With the applications of high precision operational amplifiers and other electronic components, a program-controlled electronic load was developed,so that the short-circuit currents and open-circuit voltages of the solar cells can be measured accurately. Furthermore,the multi-channel electronic switches, relays and precision resistors were combined to realize an adaptive test,and the photoelectric characteristic parameters of different thin film and module solar cells were measured accurately. The experiment results show that the test system can complete the adaptive test while the short-circuit current is in the range of 3 mA to 5.6 A and the open-circuit voltage is from 0 V to 50 V. The relative error is less than 0.6% for the measurement of open-circuit voltage and 0.8% for the measurement of short-circuit current,respectively. This result satisfies the demands of solar cell test.

To further improve the performance and stability of the primary mode for a Silicon Micro Machined gyroscope, a new driving circuit based on the self-oscillation principle and controlled by Field Programming Gate Array(FPGA) digital signal processing was proposed. Based on the characters of the digital signal processing and the principle of self-oscillation, the models for the loop phase control and oscillation amplitude control were designed and analyzed,then the control methods in which the loop phase was controlled by ‘frequency measure and adjust program’ and the signal amplitude was controlled by ‘PID control program’ were presented. Experimental results indicate that the precision of the oscillation amplitude is up to 1.5×10-5, and the driving frequency can follow the change of the primary mode resonant frequency. In addition, by introducing the digital signal control system, the thermal sensitivity of oscillation amplitude has improved from 7.69×10-5/℃ to 1.183×10-5/℃. Compared with an analog circuit, this system has more precision amplitude and better temperature adaptation.

A ground-based physical TDI CCD imaging simulation system is developed, which consists of a high precision satellite 3-axis attitude control simulation subsystem based on a 3-axis air bearing table, and a TDI CCD imaging simulation subsystem whose TDI image procedure is simulated by software using an area-array CCD. The designed system realizes the simulation of satellite 3-axis attitude control with a maximum pointing accuracy of 0.1° and a attitude steady accuracy of 0.01(°)/s, and implements an imaging simulation of 4-stage to 16-stage TDI CCD with the integration time of 0.1 s. Furthermore,the effect of satllite 3-axis attitude on the TDICCD photography is researched and the correction of the modeling theory for a space TDI CCD image motion model is proved. Experiment results show that the physical experiment results are in accordance with those of mathematical simulation and theoretical analysis, which justifies the feasibility of system design and the validity of image motion theory.

A microfluidic based Interstitial Fluid (ISF) transdermal extraction tool made from polydimethylsiloxane (PDMS) was designed to the application of continuous glucose monitoring. Four layers of PDMS fabricated by replica molding process were bonded together using oxygen plasma to construct the device, including a Venturi tube for vacuum generation, chambers for the introduction of ISF and normal saline solution, pneumatic valves for fluid control, and interconnected microchannels. Then, the output vacuum of Venturi tube and the close pressure of pneumatic valve were measured. Finally, the ISF extraction and collection function was tested. Results show that the vacuum pressure is used for fluid manipulation and a 92 kPa (absolute pressure) vacuum is achieved, when 220 kPa (absolute pressure) external pressure is applied to the Venturi tube. Moreover,the normally open pneumatic valves can be closed under the operating pressure less than 65 kPa (relative pressure). Under the control of pneumatic valves, the normal saline injection, ISF extraction and collection function of the tool are implemented by using the vacuum generated from the Venturi tube.

In order to realize the automatic image registration for infrared images and visible images, an image registration method based on the Expectation Maximum(EM) iteration of the log-likelihood function is proposed. This method utilizes the image edge as the registration point, and thus the image registration is transferred into an edge point set registration. The point set is modeled as Gauss Mixture Model (GMM), and the likelihood function of the point set registration is obtained. To solve the affine transform parameter, the likelihood function is maximized with EM iterations. And during the EM iterations, the probability density of edge point is segmented with an adaptive threshold to eliminate the outer points, and the interference of outer point with the likelihood function is overcome and the affine transform parameter is determined accurately. The experiments on image registration for infrared images and visible images are verified,and the results indicate that the proposed method is effective.

To get accurate parameters in both static and dynamic models, an identification method of model parameters based on the analysis of steady state error was presented. Firstly, the relationship between steady state errors and input and interference signals in a servo system was determined, and the influence of torque ripples on the parameter identification was eliminated based on the study above. Then, the friction torque was deduced by using the steady state error, and the static and dynamic parameters were identified through the genetic algorithm. Finally, the friction torque was compensated according to the identified model, and the compensated effect was analyzed. The experimental results show that the steady state error of uniform motion has decreaseed from 36 μm to ±3 μm, that of uniformly accelerated motion decreased from 34 μm to ±3 μm, and that of sinusoidal motion decreased from 35 μm to ±7 μm. It concludes that the dynamic and static parameters of LuGre can be precisely obtained by this identification method, and the tracking accuracy of the servo system can also be improved through friction compensation on the basis of the proposed model.

This paper focuses on the construction and optimization of neighbour graph and proposes a Supervised Graph-optimized Locality Preserving Projections (SGoLPP) method for facial feature extraction. Different from the Locality Preserving Projections(LPP) that it predefines the weight matrix and solves the projection matrix by one step optimization,the SGoLPP incorporates the weight matrix into the objective function as a learning term, and optimizes the weight matrix and projection matrix simultaneously. Meanwhile, the label information is utilized to update the weights corresponding to sample pairs in the same class and to avoid the interferences from samples not in the same class. Experiments on the Wine database of UCI show that the SGoLPP achieves better cluster performance with less iterations. For face recognition, the average recognition accuracies of SGoLPP on Yale, UMIST and CMU PIE face databases are 26.6%, 4.8% and 8.8% higher than those of LPP, Supervised Locality Preserving Projections(SLPP) and Graph-optimized Locality Preserving Projections(GoLPP), respectively, which verifies the effectiveness and superiority of the proposed method.

For the problems of complex structure, high power consumption, large size and poor anti-interference in the traditional DSP+FPGA design, a real-time multi-target tracking system based on a System on Programmble Chip(SOPC) was established and a software optimization method based on NIOS II core and its data link was proposed. The optimization technologies for algorithm structure, data link configuration, code optimization, custom instruction and C2H were researched. Then, these technologies were applied to the multi-target tracking SOPC system to optimize algorithm structure, reduce the size of code，remove the correlation of code and implement link-layer hardware conversion to enhance the performance of software pipelining. Experimental results indicate that these optimizing technologies can improve the efficiency of software by 9-18 times. The computing time of multi-target tracking is 1.3 ms/frame and the navigation operation time is 470 μs/cycle for the image sequence of 256 pixel×256 pixel in 100 frame/s. The system can realize real-time processing with low hardware costs, simple structure and stable performance.

A center-extracted algorithm by combining the Kalman filter with the Hessian matrix was proposed for the implementation of real-time stripe processing in rail wear dynamic measurement. Firstly, the image region of a laser stripe in rail wear dynamic measurement was predicted with the Kalman filter method. Then,the points with max gray were searched in the predicted region of laser stripe line by line, and they were considered as the initial positions of laser stripe image center. Sub-pixel positions of laser stripe image center were then extracted through the calculation of Hessian matrix in the initial positions of laser stripe image center. Finally, the sub-pixel image center of laser stripe was obtained in the region of stripe. The proposed algorithm reduces the searching area and the number of Gaussian convolutions greatly, and improves the robustness and speed in the extraction of laser stripe centre in the rail wear dynamic measurement. Experimental results show that it takes 1.6 ms to process every frame, and the extraction accuracy of laser stripe is also guaranteed.

In order to improve the inspection accuracy and reduce misjudgment rate for the solder joint quality inspection in an automatic optical inspection system, a pseudo solder joint inspection method for lead-free solder joints in a Printed Circuit Board(PCB) was proposed based on analyzing the solder joint image acquired by a 3-color LED structure illuminator and a color 3-CCD camera. Three steps were used to detect the pseudo solder joints. Pseudo solder joints were inspected firstly by the center of gravity method after positioning the solder joints and its components. Then, the regions of solder joints and components were figured out based on their edge, and the area method was used to inspect pseudo solder joints. Finally,the missed pseudo solder joints were inspected by the proposed color grad method. The experiment results show that the proposed method composed of the center of gravity method, area method and the color grad method can inspect the pseudo solders effectively, and the inspection accuracy can reach 99.2%.

For the large computation and the real-time tracking to be hard to achieve by traditional fragment based tracking algorithm, an improved fragment based algorithm was proposed. Firstly, in order to reduce the negative effect yielded by the background noise, the region of the tracked object was divided into more fragments. Then, the position and the range of the search region were identified according to the movement information of the object. By utilizing a fast hierarchical adaptive search approach which adopts different search patterns in different steps, most of the calculations for invalid positions were skipped, and the coordinate where the candidate was most similar with the object template was obtained quickly. Moreover, the improved algorithm was implemented and optimized on a DSP. Experimental results indicate that the improved algorithm can process the image of 768 pixel×576 pixel on DM642 at a processing speed of 30 frame/s. Compared with the traditional fragment based tracking algorithm,it shows a more precise tracking and saves the processing time about 47.5%. These results show that the improved algorithm overcome the shortcomes from traditional fragment based tracking algorithms, and can achieve a real-time tracking with better tracking performance.