On the basis of characteristics of the VUV light source at National Synchrotron Radiation Laboratory(NSRL), a V-coupling-blade Beam Position Monitor (V-BPM) is developed to detect the stability of the low energy beam radiated from an undulator. The monitor is composed of two pairs of V-coupling-blades made of molybdenum. To increase the photoemission currents and improve the response sensitivity of the probe, both V-coupling-blades are inserted in the beam edges at an angle of 10° in horizontal and offset each other along the beam direction, and the bias of ~300 V is adapted to collect the photon-electronics more effectively. It describes the principles and structure features of the V-BPM and discusses the results of performance tests. A V-BPM is mounted on the undulator beamline at the NSRL to monitor the stability of the light source and a series of data on the tests are obtained. The results show that the V-BPM can test the radiation beam position from the new undulator in the improved NSRL in real time and obtained data meet the requirements of monitoring photon-beam stability.

To solve the problem of poor roughnesses for heavy curved freeform surfaces in prism processing, a Head Mount Display (HMD) prism with decentered aspheric surfaces was developed. All the optical surfaces of the prism were rotationally symmetrical aspherics, which showed a better processing property as compared with freeform surfaces. A short focal length and a large field of view were obtained by controlling the surface curvatures and tilt angles, and the aberration was corrected by optimizing aspheric coefficients. Experiments show that the HMD prism has an exit pupil of 6 mm and a field of view of 55°. At the frequency of 30 lp/mm, the axial Modulation Transfer Function(MTF) is over 0.2, the off-axial MTF is over 0.1, and the maximum distortion is -5.37%. The test and experiment on the prism demonstrate that the surface roughness has been greatly improved to a lower level for the aspheric surfaces, which avoids the image blurring and improves the performance of visualization.

In order to measure the space angle of two non-uniplanar lines with a distance of several even tens of meters, the design proposal of a laser system for measuring the large-scale space angle formed by non-uniplanar lines was proposed. Firstly, in consideration of the existence forms of space angles formed by common targets, corresponding orientation methods were designed and a common reference in the space angle measurement was proposed. The single beam of linear structured light was used as the common reference when the distance of the targets was only large in the direction of common perpendicular. Then, the design process of a system framework applied in a specific project was analyzed. Finally, the inspection method of spatial relationship between the common reference and targets was proposed. With digital image processing, the angle between common reference and targets was measured, and the measurement result was acquired indirectly. In measurements for 30 times, the repeatability of the system is ±0.020 85°, which proves that the measurement system can satisfy the requirements of practical projects and the proposed principle is reasonable.

A kind of beam automatic alignment method used for double path amplification in an electron pumped excimer laser system was presented to ensure the beams from the amplifiers to transfer in a designed direction and irradiate on a target stably and accurately. As excimer laser amplifiers had no natural alignment references, an image-relaying structure with two cross-hairs was proposed, in which the cross-hair on the input beam was regarded as the near-field reference while the cross-hair of output beam was regarded as the far-field reference. Then, an alignment beam path was designed and the software was compiled. Furthermore, the automatic alignment experiment of double paths in electron pumped excimer laser amplifier was performed and the correlative influence factors and the alignment precision were analysed. Experimental results indicate that the automatic alignment system can achieve the beam transmission given in short time, the accuracy of alignment system is 0.63 μrad and the beam maximum restoration error is 13.75 μm. Obtained results satisfy the demand of beam alignment.

A measuring method for the refractive index of micro-quantity liquid within a glass capillary was introduced. The capillary filled with liquid samples was used as a sphero-cylindrical lens in the system and a couple of objective lenses together with the glass capillary was used to image a micro-ruler onto a CCD imaging plane. As the image size and lateral magnification were very sensitive to the liquid filled in the capillary, the refractive index of the liquid could be exactly calculated by measuring the image-lateral magnification. The refractive indexes of four standard liquid media, pure water, ethanol, ethylene glycol and glycerin were measured, and experiments show that the accuracy of this equipment is the same as that of an Abbe refractometer(±0.000 2), and the quantity of liquid samples required by measurement is less than 0.002 mL. Moreover, the refraction indexes of a series of ethylene glycol with various concentrations were also calculated by this method and the results were compared with the values measured by the Abbe refractometer. These results demonstrate that the proposed method is characterized by micro-quantity sample required, easy operation and high measurement accuracy and it is suitable for the measurement of micro-quantity liquid quickly and exactly.

By taking an xenon lamp and Light Emitting Diodes(LEDs) as luminescence media, an integrating sphere solar spectrum simulator was proposed to improve the matching precision of solar spectra. The Simulated Annealing (SA) algorithm was introduced to study spectrum matching technique, and the mechanical structures and control systems of LED module and xenon lamp module were designed. The LEDs were positioned in an annular manner, driven by a constant current and controlled by a multi-computer communication system. A cut-off filter and a programmable attenuator were mounted in the xenon lamp module, a filter was used to cut off the infrared spectrum of xenon lamp and an attenuator was used to vary the incidence radiant flux of xenon lamp entered in the integrating sphere. The simulation experiment indicates that 53 Epitex LEDs and the xenon lamp could simulate the AM 1.5 standard solar spectra successfully, and the average relative error is 5.67%. The mismatching error, irradiance uniformity and the irradiance instability were analyzed and discussed, and they all has the potential of performing better than ±3%.

In consideration of the development trends of backlights, a global dimming algorithm for side-lit backlights was proposed to solve the problems of Liquid Crystal Devices(LCDs) in low contrast ratio and high power consumption. The backlight dimming level was decided by the average luminance and the difference between the average and max grayscales of all subpixels. Then, the S-shaped function was involved to improve the static contrast ratio, by which the pixels with the grayscales more than the turning point of the curve could be become brighter and those less than the turning point could become darker, so that the static contrast ratio was improved efficiently. The algorithm was simulated and a prototype was developed to evaluate the effects of the dimming algorithm. Experimental results indicate that the power consumption of the LCD with the global dimming heavily depends on what to be displayed on the images. For 40 different kinds of images, the average power has reduced by 9.39% and the static contrast ratio improved by 124.7%. The side-lit backlights performing the proposed global dimming algorithm have the advantages of low-cost, small thickness, low power consumption and high constrast ratio.

According to the evaluation indexes of road lighting, the light distribution of Light Emitting Diode(LED) luminaries was optimized. A mathematical model of the illumination distribution for LED luminaries was established based on the relationship between the evaluation of road lighting and the illumination distribution. By taking the evaluation indexes of road lighting as the optimization conditions and the ratio(Q) of average luminance to average illuminance of road surface as an optimization target,the light distribution was obtained by solving a constrained optimization problem. Finally, the optimized light distribution was induced into lighting simulation software to simulate the LED luminaries for an asphalt trunk road. The result shows that the light distribution is asymmetric in the lateral direction, and the maximum intensity direction leans to the traffic lane. The light distribution saves 30.3% in energy consumption compared to traditional symmetrical light distribution. Moreover, the Q value is 0.072 7, which is higher than that of the normal asphalt trunk road. The evaluation of the road lighting meets the standard of CJJ 45-2006, and demonstrates that the lighting parameters are optimized well.

A measuring system based on optical coherence tomography was established to measure the distance of optical surfaces non-contactly. The system was developed by modifying the sample scanning device based on a time-domain optical coherence tomographic system. The tomographic images of optical surfaces in different depths were acquired by moving the fiber collimator in sample arm through a high precision guide. The distance of an air gap sample and a lens were measured by using this system. Results show that the measured value of air gap sample is 6.026 mm while it is 6.02 mm by using a vernier caliper. Furthermore, the measured value of air gap in the lens is 10.750 mm while its design value is (10.7±0.03) mm.The systematic error is 3.871 μm and the sensitivity is 10.5 μm. The measurement method is characterized by non-contact, high precision, and high sensitivity, and has great potential in optical material testing.

A new fuzzy 3-partition entropy approach based on a fast recursive genetic algorithm was proposed to reduce the repeated computations and to improve the processing efficiency in extraction of FISH-labelled (Fluorescence In Situ Hybridization) genes. An iteration validation method was presented to determine the window width of the membership functions and the membership functions considering the boundary conditions and gray weights were selected to perform the fuzzy 3-partition. To improve the efficiency of selecting optimal thresholds, a recursive algorithm was presented to convert the computation of fuzzy entropy to a recursive process. Then, the no-repetitive results of the processing moments were stored for the succeeding genetic algorithm to compute the fitness of each individual. Finally, the optimal thresholds were searched by the genetic algorithm in a high speed. The result of the proposed algorithm was compared to those of the several common algorithms and the classification probability and run time were analyzed as the test criterion of optimal thresholds. By evaluating various types of simulated images and real FISH images, it shows that the run time of the proposed algorithm is 1% that of other common algorithms and the misclassification error is less than 6.00×10-2. These results demonstrate that the proposed algorithm is effective for improving the precision and efficiency of extracting FISH-labelled genes.

As some aircrafts with large sizes, high accuracy or special structures need multi-instrument combination measurement networks to complete its digital measurement, this paper researched the optimal arrangement design of measurement networks. The traditional aircraft leveling theory was introduced firstly. A aircraft digital leveling algorithm was put forward, and its uncertainty model was established. Then, the effect of the accuracy of level benchmark points and the location of level points on the measuring accuracy was studied through the uncertainty analysis and error dependence simulation. Based on the measurement characteristics of the aircraft digital measurement network consisted of a laser radar and an iGPS, the optimal multi-station arrangement design of aircraft digital measurement network was put forward. Experimental results of aircraft leveling measurement show that the measurement accuracy of optimal multi-station arrangement design is 0.402 mm, which has improved by 20%. The optimal arrangement design is feasibility and effectiveness, and propsed analysis method provides reference basis for researching the multi-station arrangement design for aircraft digital measurement efficiencyth and accurately.

This paper introduces the working principles of Line-Matrix CCD( LMCCD) mapping cameras firstly and points out that the key step to realize the LMCCD camera is to develop focal plane modules. Some pivotal techniques used in the reality process of a LMCCD mapping space camera are introduced, such as stitching line and matrix CCDs together on an image base board in high precision, designing elements in high density and very low noise, controlling dark noises related to the temperature and the heat produced by the CCD under a vacuum environment, assembling the circuit board and the focal plane and soldering the CCD. Finally, it gives developing and experimental results of the LMCCD mapping camera. The results show that the stitching precision of line and matrix CCDs in coplanarity is better than 5 μm, and both the amounts of moving and parallelism are better than 2 μm. The Signal to Noise ratio(SNR) tested in laboratory is better than 90 under a typical working condition(the solar incidence angle is 30° and the reflectivity of the earths surface is 0.3). Furthermore, the temperature of the focal plane module can be controlled under 30℃ in a working period of 15 min. All of these results satisfy the requirements of the LMCCD for the CCD stitching, temperature control and the SNR.

To eliminate the out-of plane actuation of a SU-8 electrothermal micro-actuator, a novel SU-8 chevron electrothermal micro-actuator with a copper/SU-8/copper symmetrical structure was proposed. The mechanical model of the micro-actuator which contains the stiffness of a driven structure was established by using the stiffness matrix method. The geometric parameters of the micro-actuator were designed for a microgripper by using the established model. The designed micro-actuator was simulated by Ansys software and the simulation results verified the reasonability of the mechanical model. Then, the micro-actuator and the microgripper were fabricated by a simple process followed with a performance test. The test results show that when the temperature of the micro-actuator only increases 32.93℃ in a driving voltage of 150 mV, the output displacement is 2.5 μm and jaw gap change is 126 μm. During the operation processes, the out-of plane displacement of the tip is always lower than 500 nm. Finally, the micro-manipulation sequence of a SU-8 micro-part with a size of 1.4 mm×135 μm×50 μm was carried out to verify the driving ability of the micro-actuator.

An azimuth shafting bearing structure was designed for a 2 m telescope engineering project. The typical azimuth bearing structures in large Alt-azimuth telescopes were compared, and the shafting bearing scheme for integrating a high-angular contact thrust ball bearing with a radial ball bearing was put forward. The corresponding parameters of the high-angular contact thrust ball bearing were devised. Based on the Hertz contact theory and the Finite Element Method(FEM) software ANSYS, the static performance parameters within the range of 60-85° original contact angle were calculated and validated in nonlinear simulation. The 85° high-angular thrust ball bearing structure was designed with the processing technique. A prototyping bearing with a diameter of 1 500 mm was developed firstly at home and it shows good performance in a axial run-out of 0.009 mm, a radial run-out of 0.006 mm, the maximum starting torque without load of 30 N·m, and the load capacity more than 30 t. The design can offer a reliable guidance and advanced technique paths for developing the high precision azimuth shaftings of large telescope systems.

To control of a Fast Steering Mirror(FSM) stably and precisely in the circumstances of vehicles and ships, a high precision device for angle displacement measurement was designed to provide the mirror sway angle information as a feedback signal to realize the servo closed loop control. As traditional detectors for the FSM are disadvantageous to high-precision work in the mobile circumstance, a precision grating based on the Moire fringe counting measurement principle was designed by using the Chinese characters Tian splitting phase method. At the same time, the absolute zero point signal against jamming and temperature changing was designed through computer simulation and its contrast ratio was 0.25. Furthermore, in order to realize the miniature, the signal processing circuit was highly integrated. Experimental results indicate that the resolution of the measurement device is 0.15″and the angle precision is higher than 0.4″. This device satisfies the requirements of angle measurement precision in the mobile condition for the high-energy laser projecting systems.

To investigate the influence of Magneto Optical(MO) properties of a prism on systematic optical response, the MO spectra of a single BK7 prism under the condition of total internal reflection was measured, and the spectra of Kretschmann configuration composed of the BK7 prism and a gold film was measured also. Then, based on the framework of traditional optical theory, the reasons of these spectra were analyzed explicitly. The theoretical results show that under conditions of total internal reflection, the Kerr rotation of the single BK7 prism increases significantly before the system gets in the critical angle, and the phenomenon comes from the different reflection coefficients between p and s waves. When the BK7 prism with the gold film is combined to form a standard Kretschmann configuration, a huge enhancement of Kerr rotation appeares at the resonance angle due to the MO capability of BK7 material and the conversion of incident p-polarized wave to surface plasmons. The maximal experimental polar Kerr rotation of the gold film (26 nm) is 1.7°, which overcomes the disadvantage of poor response in the traditional way.

A parallel mechanism was applied to the mechanical leg of a six-legged robot to expand the application areas of the six-legged robot. A new mechanical leg based on the (U+UPR)P+UPS parallel mechanism was proposed, and the structure parameters of the mechanical leg were designed. Firstly, the kinematics of leg mechanism was discussed, and the inverse position and velocity mapping equations were presented. Then, the workspace of the leg mechanism was discussed and a 3D map of the workspace was drawn. The performance evaluation index of workspace was defined and the relationship curve of the structure parameters and the evaluation index of workspace was given. Furthermore, the kinematics dexterity of the leg mechanism was analyzed, and the 3D map for condition numbers of Jacobian matrix was drawn. The performance evaluation index of kinematics dexterity was defined, and the relationship curve of the structure parameters and the index of kinematics dexterity were given. Finally, based on the indexes of workspace and kinematic dexterity, the structure parameters were analyzed by Monte Carlo method. A set of excellent structural parameters were proposed, and a 3-DOF mechanical leg was designed. These results lay the theoretical foundation for the further study of six-legged robots.

For the problems in the design of a flat part micro-assembly system, an orthogonal optical alignment mechanism was proposed to realize high-precision assembly of the miniature flat class structure by a man-machine cooperation micro-assembly system, and the error of high-precision alignment mechanism module was analyzed and calculated. Then, a micro-assembly system platform based on the microscopic machine vision and orthogonal optical alignment was established, and the micro-assembly experiments by proposed method were carried out. The optical alignment method proposed was effectively used in the assembly of integrated and complex micro heterogeneous electromechanical system such as the flat structures of silicon micro-MEMS devices and non-silicon MEMS devices. The results show that the consistency and efficiency of the micro-assembly are improved and enhanced greatly, and the designed platform has good openness and portability. When the prism orthogonal alignment mechanism produces the error of 0.001°, the theoretical deviation is less than 0.98 μm and the assembly accuracy of the actual experimental micro-assembly platform is less than 5 μm.The precision meets the general assembling accuracy requirements of the miniature flat class structure.

The toughening effect of SiC particles was applied to inhibition of the cracks in Al2O3 ceramic for Laser Direct Forming (LDF). The experiments of coaxial laser cladding Al2O3 ceramic added SiC particles were performed on the Ti-6Al-4V alloy substrate to investigate the feasibility of Al2O3+SiC composite ceramic by LDF and explore the factors effecting on the crack susceptivity. The cracks, microstructures and the combining of the two compositions were observed by an optical microscopy, and the phase transitions were detected by X-ray diffraction. The results show that some complete SiC particles contained in the Al2O3 ceramic can inhibit the cracks, and the Al2O3+SiC composite ceramic without serious chemical reactions or combining defects can be prepared. The crack susceptivity is significantly affected by the proportion of SiC (f), laser power (P), scanning velocity (v) and the powder flow rate (n). Finally, a thin-wall Al2O3+SiC composite ceramic block(17 mm×6 mm×2 mm)without the obvious defects was formed with the parameters of f,P, v and n in 10%(weight proportion), 186 W, 300 mm/min and 1.78 g/min, respectively.

According to the stitching structure of CCDs in an Aerial Push-broom Remote-sensing Camera (APRC), an auto-focus method by using the image from the CCD overlapped sections was proposed. Firstly, the structure of CCD in the APRC was introduced. Based on the pushing-broom imaging characteristics, a Field Programming Gate Array (FPGA) which can process data in parallel model was used to receive the image data from multi-CCD overlapped sections and calculate the evaluation values of image quality at the same time. Then, three kinds of the CCD overlapped sections with the biggest evaluation values of image quality were selected as the gist for determining the focus direction. Finally, the feasibility of the method proposed was validated by an experiment. In the experiment, when the push-broom frequency is 10 kHz, the time for one focus step is 0.2 s, which illuminates that the proposed focusing method has good real-time ability, and it can offer a feasible gist for the on-track focus in the APRC.

A flotation method involving the control of the unbalance orientation of a rotor is put forward to improve the measurement precision of mass imbalance of an incomplete spherical superconducting rotor. By adding additional weights within the rotor to change its quantity of the mass unbalance, this method limits the azimuth of the mass unbalance to an accurate measurable range. As the additional weights change, the observed values change simultaneously, and the mass unbalance can be derived by a relevant formula. This method can improve the measured precision without any damage to the rotor or influence on its electromagnetic and superconducting properties. The fluctuation of the measured offset is within ±6 μm, and its Z-axis component is within ±8 μm. Moreover, it can improve the working performance of some types of the devices based on the rotor greatly, and can offer some references for other balanced measurements of the machines without mechanical support for its rotor. The measurement equipment of this method is simple, precise, and its experimental period is short.

On the basis of principles of vibrating piezoelectric vibrators to syntonize the connected diaphragms, a piezoelectric resonant air pump with a magnetic spring was proposed to improve the driving ability of the piezoelectric pump for gas moving. First, the working principle of the piezoelectric resonant air pump with magnetic spring was analysed, a dynamic model of the pump resonance was established and the main factors effecting the membrane amplitude were obtained.Then, a prototype was designed and the resonance frequency of the system and the displacement magnification of the piezoelectric vibrator were measured by an impedance analyzer and a photovoltaic displacement sensor, respectively. Finally, a testing facility to measure the volume and output pressure of the resonance pump was developed and the effects of the axial distance of magnetic spring on the output volume and pressure were analyzed. Experimental results indicate that the amplification factor is 4.5 when the sinusoidal AC driving voltage, resonant frequency and the magnetic force of the axial spacing of the magnetic spring are 200 V, 133 Hz and 9 mm,respectively. Moreover, the maximum flow rate is 524 ml/min and the maximum pressure is 9.2 kPa. Results demonstrate that the gas transportation capacity has been improved by using the proposed piezoelectric resonant air pump.

In order to improve the injection efficiency of liquid micro injection and the injection directivity of powder micro injection, a borosilicate capillary and a quartz glass tube are chosen as raw materials to design and manufacture the in-line combined micro nozzle and the coaxial combined micro nozzle based on stable drawing and forging process. Using these tow combined micro nozzles, liquid and powder micro injection experiments are performed on a micro injection experiment system based on the digitalization of microfluidic technology. In liquid micro injection experiment, the regular and uniform droplet array with a average diameter of 180 μm and no satellite droplets is prepared with one injection in a 4×2 in-line combined micro nozzle, by which the efficiency of micro injection under in-line combined micro nozzle is improved as compared with that of single micro nozzle. In powder micro injection experiment, the injection angle is reduced from 33° to 10° and the width of formed powder line is decreased from 450 μm to 300 μm as compared with that of single micro nozzle. The results indicate that the powder flow of the major micro nozzle in coaxial combined micro nozzle is constrained with the assisted micro nozzle and the injection directivity is improved greatly.

A lightning identification system that can realize frame-frame background removal algorithm is designed to accomplish the lightning identification and coding in real-time on a meteorological satellite. This system divides the high-speed image captured by a lightning imaging sensor into eight regions, and the lightning detection of each region is completed by an identical subsystem respectively. Each subsystem uses seven Static Random Access Memories(SRAMs) to read and write data in the pipelined architecture to implement the background evaluation and realize frame-frame background removal algorithm. Experimental results show that this system can receive high-speed data flow up to 2.88 Gbps, realize background signal evaluation and removal, and identify lightning events in real-time. These results show this system is able to identify the lightning on meteorology satellites.

A gain scheduling three-loop autopilot was set up to realize the real-time change flight trajectory of a missile, and its structure characteristics, feedback parameter acquiring, control performance and the transfer flight ability were researched. With the principle of state feedback and the pole assignment, a missile longitudinal flight control mathematical model was established, then the influence of the missile velocity, dynamic pressure and the elevator efficiency on the control instruction were analyzed in a typical three-loop autopilot. Combing with gain scheduling theory, the variables of velocity ratio, dynamic pressure ratio and elevator efficiency ratio were induced to correct the system nonlinearity, and the method to get feedback parameters was derived. In the end, three-loop robust stability was also analyzed. The comparison of the calculation results and the flight test data shows that the measuring accuracy of atmospheric density and atmospheric temperature is better than 1.7% and 3.0%, respectively, and the semi-physical simulation experiments prove that when the velocity changes to 20.6% in the variable trajectory, the cut-off frequency, overload steady-state gain and the phase margin are changed from 41.78 rad/s, 0.87, 78.19° to 41.71 rad/s, 0.88 and 72.5°, respectively. Moreover, the phase margin is also larger 71.46° than the minimum of basic trajectory, and the autopilot still keeps a good control characteristic when the dynamic pressure changes 77.6%. These results prove that the gain scheduling three-loop autopilot meets the requirement of missiles for real-time transfer trajectory flight control.

A design method for ultra high resolution image systems based on full frame CCD KAF50100 was proposed, in which a mechanical rolling shutter was adopted to control the exposure. The Analog Front End (AFE) chip AD9845B was used to process and digitize the analog output of KAF50100 and the digital image signal from the AFE was buffered and arranged in a block ram of Field Programming Gate Arry(FPGA), then was sent out through a Low Voltage Differential Signaling(LVDS) interface. All the control signals and drive timing signals were generated by the FPGA and the RS422 interface was used to receive command and upload the status of system. The least square method was adopted to correct the non-uniformity from four analog outputs of the CCD sensor. Experiment shows that the CCD sensor KAF50100 can work at a top output speed, its pixel output rate is 4×18 MHz, and the maximum frame rate, output noise and the dynamic range are 1 frame/s, 2.76@12bit and 63.4 dB, respectively. With better flexibility and extendibility, this design method can be widely used in ultra high resolution imaging fields like visible light underwater detection, satellite remote sensing and astronomical observation.

A method based on Bayer filtering to obtain color images was introduced, then the Modulation Transfer Function (MTF) of a color area Charge Coupled Device (CCD) which uses the Bayer filtering to obtain color images was established based on the discrete sampling characteristic of CCD. The MTF values for the color area CCD using Bayer filtering were compared with those of a monochrome area CCD. Obtained results indicate that the MTF value for color area CCD using Bayer filtering is notably lower than that for the monochrome area CCD. When the side length of a pixel is equal to the corresponding center distance, the MTF value at Nyquist frequency is 0.067 6 for the color area CCD using Bayer filtering while that is 0.405 3 for the monochrome area CCD. Moreover, in an imaging experiment using a resolution card, the resolution of the former is lower about 40% than that of the latter in the same imaging condition. These results are coincident with comparative analysis results of MTFs in general.

This paper investigated the working scheme of a spaceborn signal processor applied to the USA Midcourse Space Experiment (MSX) for target detection and tracking. The moving target indicator in the MSX performed the trace extraction in two stages, namely, “nominator-confirmer” procedure. the first stage used a binary matched filter to “nominate” streak paths which looked like they might actually contain signals and the second one did a energy accumulation decision to “confirm” the nomination of the first stage. Based on the study of above procedure, an improved method was proposed to detect moving targets. In “nominator” stage, the velocity restraint condition was introduced to binary matched filter for reducing the probability of false alarm. In “confirmer” stage, the criteria of “minimum sample mean difference” were added to that of “maximum sample mean” for improving the probability of detection. The improved method could be widely used in complex background images. Finally, an experimental method was designed to analyze the algorithm performance of two different methods. The experimental results show that the number of streak paths filtered by improved method is lower than 50% compared with that former method, and number of clock cycles is reduced from 8.0×106 to 7.1×106. Improved method can effectively detect multi-moving targets in real time by expanding the reasonable range of threshold to nearly 20.

Atmospheric transmittance is an important factor affecting the radiation measurement precision. Generally, the error of conventional atmospheric transmittance calculated from atmosphere parameters by using a atmosphere model calculation software such as MODTRAN is above 15%, which limits the target radiation measurement precision greatly. Therefore, a new radiation measurement method was proposed, by which the ratio between the really-measured atmospheric transmittance and the calculated one for a distance was taken to obtain an atmospheric transmittance correction coefficient. Then, the correction coefficient was used to correct atmospheric transmittances of other measurement distances. Experiments show that the precision of the corrected atmospheric transmittance is better than 8.1%,which improves the target radiation measurement precision greatly. A radiation measurement experiment was performed by using a midwave infrared camera and a blackbody, and the results show that the radiation inversion precision obtained by the conventional method is about 20% and the precision obtained by the proposed method using the corrected atmospheric transmittance has been better than 10%.

A new three-dimensional deformation measurement method was proposed based on digital image correlation and binocular stereovision to measure the full-field deformation in metal sheet welding processes. Firstly, an algorithm based on a seed point was proposed to provide a reliable initial value for the least-square nonlinear optimization in the correlation matching. Then, the algorithms for three-dimensional coordinate reconstruction, three-dimensional displacement and the strain calculation were discussed separately. Finally, based on these algorithms and the VC++6.0 platform, a welding deformation measurement system was developed for the full-field deformation of metal sheet welding. With self-developed image acquisition devices and a standard material testing machine, a steel standard tensile test and a welding deformation measurement test were conducted to validate the performance of proposed mthod. Experimental results demonstrate that the accuracy of strain measurement is not lower than 0.5%, which is very close to that of a extensometer and the measured 3-D displacement time history is reasonable. It concludes that the proposed method can satisfy the requirements of non-contact, higher precision, and it is a more intuitive way for full-field deformation measurement during metal sheet welding processes.

As the two edge pixels of a heavy rail image is not identical, the classical edge operators are difficult to achieve the edge extraction and segmentation. Therefore, this paper proposed a new algorithm to enhance and extract images. A strong contrast stretching algorithm was used to stretch the rail surface and the background differently, enhance the edge information and weaken the background information. Then, the maximum variance method was taken to select the optimal threshold to implement the binarilization. Finally, the recursion connected domain marker algorithm was used to locate the pixel coordinates of edge to achieve the image segmentation. 30 images were chosen to a discretion experiment, and results indicate that the gray features of image edge are enhanced clearly, surface textures and false edges are restrained availably. Moreover, the pixel width fluctuating range is reduced from -0.64% to 0.34%. With the discrete pretreatment algorithm via addressing global array of a register, the segmentation time has been decreased to 10.165 s. The algorithm is better than the classical edge operators in the precision, correctness and the timeliness and is more suitable for on-line detection systems.