A novel demodulation scheme for the fiber Bragg grating sensor based on a new-pattern polarization-maintaining fiber loop mirror is designed to achieve highly stable and accurate wavelength interference demodulation. Based on the matrix optics theory, a model is established for analyzing the elementary principle of the interference demodulation using this fiber loop mirror. Then, the influence of structure parameters on the demodulation precision is deeply analyzed, and a numerical simulation is undertaken. Meanwhile, the optical and electrical modules are developed, and the testing software and monitoring software are designed for eliminating the system structure errors based on LabVIEW. By a sensor detection experiment, the system precision on temperature in the range of 20-90 ℃ can reach the resolution of 0.03 ℃ and the accuracy of ±0.1 ℃. The experimental results agree well with theory analysis, which sufficiently shows that this system has advantages in favorable stabilization, much higher precision and stronger applicability.

A new computer controlled polishing technology called the fixed abrasive machining technology is described in this paper,and a removal function model for multi-pellet polishing pads is established based on the removal function theory of plane motion .The parameters of the model,motion deflection distance of the polishing pad and the distance between pellets and pellets are optimized by a approaching factor and a curve rms distance to use in experiments. The comparison of the theoretical model and the experimental results show that the error between the theoretical maximum removal rate and the experiment data is 0.007 3 μm/min, and its deflection ratio is 5.58% . The rms distance warp between the theoretical removal function curve and the experimental curve is 0.084 9 μm and its deflection ratio is 7.01%. The veracity of the theoretical model is verified by experimental results, which can predict the feasibility of the fixed abrasive polishing technology and can lay a foreground for the SiC mirror precision fabrication field.

In order to solve the calibration problems in line structured light sensor systems such as requirements for high accuracy calibration models, complicate calibration procedures and so on, a calibration method based on perspective-three-points was proposed. A calibration model was established by introducing a planar target that could move freely with three collinear and known mutual location points. The calibration point coordinates in a camera coordinate system could be calculated by using the imaging information of three points and the light stripes on image plane of the camera. The spatial pose between the camera and the light plane could be calibrated by moving the reference plane at several arbitrary places in the sensor range. Experiments show that relative accuracy of the sensor is about 0.72%.The method is suitable for the field calibration,for it has advantages in its simple and flexible calibration, the cost of calibration equipment, and not to calculate the transformation matrix among different coordinate systems.

A method to characterize the retardance of retardation films at different wavelengths was presented. Based on a 1/4 λ wave-plate rotation method,the system used a spectrometer as the light source,color filters and a photodetector. Some samples were measured using the system,and the characteristic spectra of the retardation films were obtained by analysing the measurement data.The obtained spectral data show that the blue light retardation is (1-3.5%) that of the green retardation and the red light retardation is (1+2.6%) that of the green one. These results indicate that the dispersion effect of retardation films are not obvious,and the retardation characteristics of different retarders generated by the same material are almost the same,which would be useful to study how the color liquid crystal display is affected by dispersion effects of retardation films. Because the variation of the light source and other errors related to light intensity are not so critical for the measurement in this system and some 1/4λ wave-plates affect on the other wavelengths is very little, the retardance can be measured in high accuracy and high repeatability. In principles,the system accuracy can measure any retardance ,and can be used in characteristic spectrum analysis of other phase retarders.

A improved technological process to the surface modification of a RB-SiC substrate mirror was proposed according to the material special properties of the large scale and complicated light weighted RB-SiC materials for space projects. Using a high energy Kaufman ion source for the Ion Beam Aided Deposition(IBAD),the substrate was pre-carbonized and coated by a carbon buffer layer. Then,a Si layer was coated upon the buffer layer for surface modification,and the properties of the sample were tested. In comparison to the process simply using a Hall ion source for the IBAD,the coating has a more compact and uniform structure and a better property in polishing. After surface modification,the surface roughness(rms) of the substrate is reduced to 0.635 nm which is the same as that of the S-SiC substrate,also the surface reflectance is obviously enhanced to that of the fine polished zerodur glass. The conclusion can be drawn that this technological process of surface modification for the RB-SiC mirror is quite reasonable and effective.

In order to obtain new wavelengths and high efficient laser outputs,an external resonator Raman experimental setup was established and the properties of YVO4 crystals were measured. Then,a high optical quality YVO4 crystal with the largest size of 28 mm×40 mm and a mass of 126 g was grown by Czochralski pulling method.The Raman output properties of YVO4 crystals in the external resonator were researched using a mode-locked Nd:YAG laser with a pulse width of 40 ps as the pump source.The results indicate that the total conversion efficiency for multi-wavelength output has reached 51.4%,and the largest output energy is 2.21 mJ. Moreover,the highest conversion efficiency for the first Stocks Raman output at 1 175 nm is 32%,and the largest output energy is 0.57 mJ. The conversion efficiency for the second Stocks Raman output at 1 313 nm is 12.3%,and the largest output energy is 0.36 mJ. These results reported here show that the external resonator structure can improve the Raman performance of YVO4 crystals evidently.

With the aim to improve the performance of a biochemical analyzer,an uniform integral time method for CCD data acquisition was presented in the paper. Different integral time data were calculated based on a Photo Diode Array(PDA) driving program and the photoelectrical transformation principle,and the linearity region of the biochemical analyzer was confirmed according to both performance parameters and experiments. Then,different integral time dark currents were recorded in the experiments and they were removed from an original spectrum to ensure the analyzer error in minimum. The experimental results show that the precision of the analyzer has been improved by 25%,and the work efficiency is 1~3 times higher than those of the original methods. It is proved that the using the proposed method can simplify the experimental procedures and can provide the higher precision for the analyzer even in sample concentration to be changed greatly.

In order to realize the high accuracy seismic exploration in high electro-magnetic fields and to improve oil producibility and oil recovery efficiency,a three-component photoelastic waveguide accelerometer was developed. Based on the photoelastic effect,a Mach-Zehnder integrated optical interferometer was presented to measure the three-orthogonal components of acceleration and a combined three-component simple harmonic vibrator was designed to reduce the cross-talk among the acceleration components. According to the phase variation of a LiNbO3 waveguide under the action of the applied acceleration,the cross-axis sensitivity and transverse sensitivity ratio(TSR) were analysed. Experimental results indicate that the natural frequency of the sensor is 3 500 Hz; response bond is from 0 Hz to 3 000 Hz,and the transverse sensitivity ratio is 0.11%,which can satisfy the sensor requirements of high accuracy seismic exploration.

A new design method with good applicability and easy to implement in technology is presented for broadband metallic diffraction gratings with single blaze surfaces. In consideration of the principles of two types of grating anomalies,Rayleigh type and resonance type,the occurrence conditions of the grating anomalies are discussed. On the basis of electromagnetic theory of gratings,the blaze characters of diffraction efficiency both for TE and TM modes are numerically analyzed,and the relationships among the grating anomalies,the spectrum bands of gratings and the blazed wavelengths is obtained. An idea of designing diffraction gratings is proposed to compensate the Rayleigh anomaly or resonance anomaly with the first blazed wavelength of TM mode. It is pointed out that the compensating effect of the first blazed wavelength of TM mode to Rayleigh anomaly is only the limitation case to the resonance anomaly. The design examples of broadband metallic gratings with different groove densities applied in UV-visible and infrared spectrometers are given.Experiments indicate that the diffraction efficiency of the broadband methallic grating has been over 40% in different spectral bands,which shows that the compensating effect method is better than traditional design methods of broadband metallic gratings,whether in theory design or technology process.

In order to improve the resolving power of a scanning holographic system,the diffraction of a Fresnel-zone-plate (FZP) functioning as a coding aperture was studied. In consideration of the lens as a phase modulator,the integral diffraction equation of the FZP was solved digitally. As compared with the distribution of the diffraction intensity with and without a lens,an equivalent distance was introduced to simplify the analyses and a virtual Fresnel diffraction plane was proposed accordingly. Meanwhile,a critical diffraction distance was defined to describe the diffration effect of the FZP quantitatively. Experimental results indicate that the critical diffraction distance is 16 cm,when the number of fringes and the radius of the FZP are 10 and 0.75 cm respectively and the focal length of lens is 7 cm. Furthermore,the critical diffraction distance increases as the focal length of lens or number of fringes of FZP increase.These results reported here show that the complicated system can be simplified due to the introduction of the equivalent distance and the diffraction effects influenced by some factors can be quantified by proposal of the critical diffraction distance,which can provide directions for the imaging system design.

An acousto-optically Q-switched CO2 laser is developed to satisfy the requirements of research fields such as laser ranging,environmental detection,spacial communications and the mechanism study of the interaction between laser and materials. In consideration of the influence of various factors on the outputs of the acousto-optically Q-switched CO2 laser,an optimization method to design the acousto-optically q-switched CO2 laser is presented by using the theoretical analysis and calculations of the laser rate equations for the main parameters of laser outputs. Then,a verified experiment is undertaken. The experimental results show that the laser can offer the laser parameters in the pulse repetition frequency (PRF) of 1 Hz-50 kHz,a pulse width of 180 ns and a peak power of 4 062 W,which agrees with the theoretical results well. The results demonstrate that a miniature laser with high repetition rates, narrow pulse widths and high peak powers can be realized by optimally choosing an acousto-optically Q-switched crystal and rationally designing an optical resonator,and it can also realize the wavelength tuning and coding output by selecting the grating lines and controlling the TTL signals.

A focusing method by adjusting two slit mark images was proposed to realize accurate focusing of a non-mydriatic fundus camera for an observed retina image in dark. Two deflection-angle prisms were used to project a slit mark on the retina,then light propagated through the slit mark in two different directions based on the bending of deflection-angle prisms. The slit mark image on retina was broken into two images and the two slit mark images would be taken apart when the fundus was out of focus. The separated distances and directions of the two slit mark images were used to decide the degree and direction of the camera focusing. The fundus camera would be in focus when the images of these two slit marks were adjusted to be in a line. Experimental results indicate that the focusing precision of the non-mydriatic fundus camera by judging the two slit mark images has increased by 50 degree (0.5 m-1) as compared with that by judging the dark retina images,which shows this focusing method has advantages in simple adjustment and high precision.

A design method of aspherical surface for a panoramic imaging system with two mirrors was proposed based on Multi-population Genetic Algorithms(MPGA).To eliminate the astigmatisms induced by mirrors and to improve the image resolution,an algebraic expression of astigmatism for panoramic imagers was induced based on the generalized Coddington equation and the theory of geometric optics. Then, an optimization process for mirror profile design was proposed to eliminate the astigmatism and to provide a purposely-designed projection formula with aid of the MPGA. A series of polynomial expressions of aspherical surfaces were obtained and procedures of the design were presented. In order to facilitate ray tracing and aberration calculation,an even asphere surface model was established by using a hybrid scheme combining the MPGA and the damped least squares. Finally,a prototype of the omnidirectional panoramic imager with f′ of -1.2 mm and F/# 1.5 for 360°(35~90°) was developed,which can provide clear panoramic ring images.

In order to provide a design tool for the photoelectronic tracking systems in remote controls,an embedded virtual dynamical signal analyzer is established,and its applied algorithms such as system identification are investigated. Based on the platform of a photoelectronic tracking system and combining the concept of a virtual device with the embedded technology,an idea for an Embedded Virtual Dynamical Signal Analyzer(EVDSA) is presented and the system infrastructure is established. Three layers of softwares for the system infrastructure is designed respectively and the application layer based on the system identification technology is introduced especielly. The current loop and the speed loop of the photoelectronic tracking system are tested by the EVDSA and the experimental results show that the photoelectronic tracking system meets the design demands for the bandwidths of a current loop of 266.05 Hz and a speed loop of 20.3 Hz.The virtual device has good accuracy and practicability,which can satisfy the design tool requirements of precise,convenient and rapid speed design levels and can improve the efficiency of the photoelectronic tracking system.

With the aim to improve measuring sensitivity,enlarge dynamic detection range and to realize the temperature compensation of microflow sensors in microfluidic systems,a novel pressure microflow sensor based on Lamb waves is set up in this paper. The sensor mainly composed of two Lamb wave sensors connecting by a microchannel. Based on the sensitive properties of the film stress in the lamb sensors and the relationship between the frequency output and the pressure in a proportion,the sensing system measures directly the frequency output of each sensor to obtain the pressure difference between two ends of the microchannel and also construct a different frequency measurement structure between two Lamb wave sensors to realize the temperature compensation. For a microchannel with 20 mm in length,1 mm in width and 50 μm in high,experimental results indicate the frequency difference between two Lamb wave sensors varies linearly with the flow,and the linear correlation coefficient is 0.999 9. Before optimization of the microflow sensor mentioned above,the minimum detectable flow is 0.627 μL/s.

An automatic-leveling system is developed.Three mechanical supportings are chosen to sustain a large-size photoelectric equipment,and a obliquity-sensor fixed on the pedestal of the photoelectric equipment is used to measure the size and direction of the obliquity angle from a baseplane. Based on a leveling algorithm,the obliquity angle is converted to elongations of the mechanical supporting to drive the mechanical supporting to level the baseplane. The reasons for choosing three supporting points and mechanical supporting are introduced,then mechanical supportings and the entire transmission system are designed.A driven circuit and automatic-leveling arithmetics are given also. Automatic-leveling experiment indicates that the leveling precision of the automatic-leveling system reach 0.003°,and the leveling time of photoelectric equipment is about 120 s,which shows that this system is better than formerly systems on prepared time and precision.

The high accuracy time interval measurement is a key technique to a pulsed laser ranging. In this paper,a new high precision time interval measurement method combining with a pulse counting method,a multi-phase sampling and a delay method is proposed to use in the multi-level time interval measurement in a Field Programming Gate Array(FPGA). The pluse counting method is used to realize the coarse counter to guarantee a large dynamic range,and a Phase Locked Loop( PLL) in the FPGA is used to generate the N clock pulses with the same frequency and uniform distribution phase as counting clocks. Based on the equal precision frequency measuring,the resolution of the time conversion is improved to Tclk/N. Moreover,a latch Flip Flops is used to form a delay chain to measure the fine time interval. Experiments show that the proposed method has solved the problem of the traditional multi-phase sampling approach,to which phase shift resolution is decreased with the frequency increasing,so that the fine resolution can be obtained by limiting delay lines and no increase of counting clocks. The test results show that the maximum measuring time of the module is 163.8 μs,the conversion time is short and the standard error of the measurement is less than 71 ps when repeated measurements are carried out,which can satisfy the requirements of higher precision systems.

In order to improve the ground pixel resolution and positioning accuracy of ground targets,the focusing control systems should be designed to compensate the defocusing of the CCD plane in a tridimentional mapping camera. According to the optical structure and focal depth of the camera,a focusing mechanism based on a cam is designed,and focusing schemes and focusing quantity are confirmed.Moreover,the reliability design and a removing return stroke error method is used to guarantee the strength and rigidity of the mechemism and to improve the focusing accuracy. Analysis of transmission error and validated tests indicate that the error of this focusing mechanism is ±1.71 μm;linearity of CCD plane is 11.2″ in the X direction and 11.8″ in the Y direction;and the circumaxis angle of CCD plane is 4.7″.The result of precision analysis shows that the focusing mechanism can meet the requirements of the users.

Extracting the center of a stripe is one of the key techniques of a structured light measurement system. An approach to increase the extracting resolution for the centers of a laser stripe and the extracting precision is investigated.Firstly,full resolution initial centers of the light stripe is presented based a threshold algorithm.Then,the gray grades of the field along the initial points of the stripe are calculated with a Sobel operator,and the normal direction at the initial points of the stripe is calculated by the Bazen method.Finally,the energy centers of the stripe are accurately gained in the normal direction.A contrast experiment in the V-shaped test block with a camera in resolution of 768×576 indicates that the resolution of the centers is 1086 dots without redundance,and the number of effective dots has increased by 42.5％ as compared with that of barycenter method.In addition,the average distance between the extracting corners and the line fitted by the extracting centers is the shortest one,only 0.294 and 0.306 pixel. The proposed method can calculate exactly the normal direction of the stripe,and computing time is 75.8 ms. Experiments show that the approach can avoid a mini density for a large curvature surface of the measured object and can get full resolution centers,it is suitable for online accurate measurements.

In order to evaluate the performance of software and instruments,the assessment mean lines of roughness profiles are researched by means of Gaussian filtration,least square mean lines and the least square conic fitting on the basis of the software standard ISO 5436-2. And by simulating manufacturing process data,measuring EDM surface data and honing surface data,the typical roughness profile parameters Ra、Rq、Rp、Rv、Rsk、Rku defined in ISO 4287 are assessed and their deviation values from standard results are analyzed. Experimental results indicate that for the simulated process data,the results given by three kinds of mean lines are all better than those of the original ones,except the deviation value of Rsk is relatively bigger by 50% or so than that given by the least square method. For the EDM surface data,the deviation values given by the Gaussian filtration are smaller and those given by other two least square mean lines are a little bigger.The relatively bigger ones are Rsk with relative errors of 3.55% and -7.45% and Rp with relative errors of -3.45% and 3.95% respectively. For the honing surface data,the parameter deviation values are all bigger computed under three mean lines because of the jumping bump,and the deviation values of parameters Ra、Rq are relatively smaller,but the others are bigger. After excluding the jumping bump,the deviation values of Ra、Rq are still relatively smaller,those of Rsk and Rp are minishied from big to small,but those of Rku、Rp are still 40% or so. It can be concluded that the roughness parameters almost have no distinct difference based on the three mean lines for the profiles without jumping bump. Therefore,for the common precision profile without distinct periodical waveness components and bigger jumping bumps,the typical roughness parameters Ra、Rq can be assessed by the least square method which is simple and easy to be realized.

With the aim to measure and study the electricity generating performance of Piezoelectric Ceramics(PEC),a measuring system for the electricity generating capacity of the PEC is developed. A piezoelectric cantilever dibrator is designed and the effect of Optimal Thickness Ratios(ORT) of the PEC to the metal plate on the piezoelectric dibrator is analyzed by Finite Element Method(FEM). The results show that the output voltages and power are maximal at the resonance frequency. The ORT of the piezoelectric cantilever dibrator is 0.5 obtained by FEM,which is consistent with the experimental results. Moreover,the output voltages of the piezoelectric cantilever increase as the resistive load increases,and the voltage approaches to 5.4 V when the resistive load is 50 kΩ. Nevertheless,the electrical power decreases when the load resistances are further increased,and the maximal output power is about 1.18 mW from the piezoelectric cantilever in a resistive load of 10 kΩ at the resonance frequency.Experiments show this power level is high enough for some wireless communication systems.

The application of the single-chamber piezoelectric pump is limited in many fields,for it has self-priming performance and bubble tolerance. It is an efficient way to adopt a multi-chamber structure to enhance the performance of piezoelectric pumps. The analysis result on changing processes of the volume and pressure in the chamber of a pump shows that the double-chamber serial piezoelectric pump only can be driven by a tandem mode,and the double-chamber parallel piezoelectric pump only can be driven by a parallel mode,respectively. The experiment results indicate that the maximum flow of the double-chamber serial pump is 1150 ml/min in a sine wave voltage of 200 V and a frequency of 152 Hz,and that of double-chamber parallel pump is 640 ml/min in a sine wave voltage of 140 V and a frequency of 220 Hz.These data reported here show that the multi-chamber with a serial structure can enhance the efficiency and performance of pumps.

In micro cutting process, cutting forces affect greatly the tool life and the machining precision. Therefore,it is very significant to study the variations of cutting forces and the factors affecting the forces for determing the machining parameters and for improving the performance of machining systems. In this paper,the theoretical formulas and mathematical model of the micro cutting are established to calculate the cutting forces by using the axi symmetric theorem and by considering effects of the rounded edge radius of a cutting tool.The effects of cutting parameters,tool materials and workpiece materials on the cutting forces are analyzed,and the experiments are coincident with the theoretieal analysis well. The results show that,when the cutting depth ap,amount of feed f and the cutting speed v range in 0.002-0.032 mm,0.01-0.20 mm/r and 20-120 m/min,respectively,the forces Fz and Fy change in 100-1030 N and 40-700 N. Also,decreasing the rounded edge radius will reduce the contact length between the tool flank surface and the workpiece and the deformation of the workpiece materials below the cutting edge so that the quality of the machined surface is improved. Furthermore,the effect of cutting speed on the cutting force can be reduced by controlling the ratio of the cutting depth to the rounded edge radius,and the ratio of Fz/Fy can be controlled by adjusting the ratios of the feed per revolution and cutting depth to the rounded edge radii.

To explore Traveling Wave Electroosmotic (TWEO) pumping microflow,a TWEO pumping model is established in a closed microchannel,and simulation and experiments are presented. Based on the TWEO driving mechanism,the mathematic models of an electric field and a flow field in the closed microchannel are established respectively to solve the electric field and flow field problems. From the flow field simulation results,the measuring velocity on 2/3 of the microchannel height in experiments is proved to be reasonable. The experiment and simulation are undertaken to determine the capacitance impedance coefficients.The results show that for KCl solution with conductivities of 1.5 mS/m,7.7 mS/m and 16.9 S/m,its capacitance impedance coefficients are 0.05,0.035 and 0.025,respectively,and conductivities of the solution are proportional to the capacitance impedance coefficients linearly. The results can offer a theoretical and experimental base for further research and applications of the TWEO.

A double-exposure digital holographic interferometry is proposed to use in the inspection of the air bubble defect in radial tires. A series of digital holograms recording the initial state and deformation state of a radial tire are recorded in a computer,then the reconstructed images are obtained by the computer. In comparison with the interference fringes between the deformed wave-front and undeformed wave-front from the defective tires and the regular tires,the characteristics of the interference fringes and phase distribution from different tires are obtained and the detection rules are established. The results show that the double-exposure digital holographic interference fringes in the air bubble defect area have special features,which indicates the proposed method can be used in tire defect inspection in high precision,and the horizontal resolution of the system is 16.4 μm.

A digital image stabilization algorithm based on a particle filter is present to remove the interframe vibration of video images. Firstly,the corners of video images are extracted using Harris operators,then the mapping relationship between the current frame and the reference frame is established and global motion vectors are obtained by computing least-square solution based on an affine transformation model. Finally,the particle filtering is used to smooth motion vectors to realize the motion compensation of video frames. Experiments are undertaken for a video sequence with 80 frames,it could be found that the average Peak Signal-to-Noise Ratio (PSNR) of the image sequence after image stabilization is 24.88 higher than that of the original video sequence while providing the image stabilization accuracy and the processing time low than 1 pixel and 30 ms respectively.These results show the proposed stabilization algorithm improves image quality effectively,and it can not only alleviate the vibration but also can preserve the initiative motion of the camera.

The extraction of the line features is very critical for traffic sign recognition. Aiming at the drawbacks of existing Hough transform,such as peak diffusion in a parameter space,heavy computational complexity and massive storage requirements,an adaptive many-to-many mapping scheme insensitive to the threshold is proposed to exactly extrac the lines from a image with low computational complexity. Experimental results reveal that the range of selectable threshold of the new method is three times that of MTO~~θ,while the running time is 10% that of MTO~~ρ. These results reported here show that proposed method is more efficient for extracting the lines in traffic sign images with acceptable running time,and it can also be used for line detection in symbol images and bill images.

An image restoration method based on pixel motion-blur character image segmentation is presented to restore a kind of space-variant blurred image taken by a moving camera. According to the convolution of motion blur,the motion characteristics of a target and a background and the shift superposition theory are analyzed. It is concluded that the gray changes of neighboring pixels within a superposition region are similar. According to the feature,the blurred edge of the target in the superposition region can be detected,and the target edge parallel to the motion blur direction is detected by a Canny operator. Then the background information of the superposition region is subtracted from the target image,and the target image is restored by adding zeros and a deconvolution algorithm. The experimental results show that the restored target image has clear textures and the improved Signal-to-Noise Ratio (SNR) is 9.64 dB,which shows proposed approach can effectively realize the restoration for this kind of space-variant motion blurred image.

In order to realize the automatic 3D measurement and reconstruction for a free-form surface,an automatic measuring system for the stereo vision is established and its adopted algorithms such as the nonlinear calibration for the measuring system and the match algorithm for stereo images are investigated. A nonlinear calibration method based on vanishing-point theorem is applied to calibrate the system and the image features of a free-form surface are enhanced by projecting arrays of dots. Then,an array coded algorithm based on direction and distance restrictions is presented to code projecting dots. After setting an original matching point and four original matching directions automatically based on the positions of red points,all other white dots viewed both in two images can be coded by the proposed method and the images in two directions can be matched automatically. The experiments of 3D measurement and reconstruction for a plaster model are undertaken,and the experimental results show that when scale is 0.08 mm/pixel,the square error of 3D measurement for dots is 0.05 mm,which can reach a sub-pixel measurement precision.Moreover,the projective points on a picture can be coded precisely and the matching between two images can be performed effectively by the array code method based on direction and distance restrictions.

With the purpose of improving the assembly performance of an Optical Burst Switching (OBS) network,the assembly modular for the edge node of the OBS is studied in this paper and a novel design scheme for Assembly Controllers (AC) is proposed. By introducing the feedback of BDP sizes,a time-threshold based modular is changed into a closed-loop system different from former assembly modules using an open-loop system. By measuring the network traffic flow,a suitable time-threshold for a timer is chosen according to the control curve of the novel modular,meanwhile,the control curve is modified according to the BDP size feedback. Simulation results show that the novel modular can improve the distribution of BDP size by 4.95%,and decrease the assembly delay by 15％ as compared with that of the time-threshold based modular. Therefore,the novel modular can not only adjust the time threshold according to the traffic flow,but also can control the distribution of BDP sizes.

The centroid and centroidal deviation of bullets are important design parameters of bullets. In order to improve the measuring accuracy and efficiency,a high accuracy inspecting system for measuring the centroid and centroidal deviation of bullets based on three point method is designed. The working principle,relevant calculation formulae and the measuring equipment for measuring the centroid and centroidal deviation of bullets are presented,and the main factors affecting measuring precision are discussed.Then,a multiple measuring platform with automatic leveling and column structure autorotation of 90° is put forward. The given experimental results indicate that the measuring precision of the centroid and the centroidal deviation is no more than 0.1 mm and 0.01 mm respectively,which satisfies the qualification needs for high accuracy measuring the centroid and centroidal deviation of bullets. The system has wide measurement ranges,low manufacture costs and broad application and development prospects.

A parallel processing system was developed to realize the real-time restoration and super-resolution processing of a huge amounts of large size images after investigating its technical scheme,working principle,key algorithms,hardware architectures and parallel technology. The working principle of the system and key algorithms of deblur,denoise,super-resolution were presented based on the technical scheme,and two kinds of the architectures,Digital Signal Processing(DSP) and Cluster systems,were analyzed and compared. It is concluded that the Cluster system is more large scale parallel processing. A parallel algorithm model based on the PPF architecture and a multi-layer parallel optimization technology based on the hybrid architecture of MPI+OpenMP were introduced.The processing effect of the proposed algorithm and the processing speed on the DSP and computer were analyzed.Following by prediction of the scale and performance of system,How to choice the number of processors was also discussed. Experimental results show that the parallel system can reduce run time from 2 700 s to 29.39 s,and can provide improved images with better sharpness,contrast and resolution.

An image restoration algorithm is proposed to restore the images simultaneously degraded by a multiple blur during aerial imaging. Based on the analysis of the physical model of the multiple blur,the multiple blur can be restored by deblurring the every sub-blur step by step. The Point Spread Functions (PSFs) of every space-invariant blur in the multiple blurred images are combined together by convolving all the linear PSFs to reduce the accumulation of calculation errors during the image restoration,so that the multiple blur made up of the space-invariant blurs can be eliminated by a deconvolution restoration. By analyzing the error noise caused by the image restoration,a Wiener filter is applied to suppress the error noise,which can increase the PSNR of the restored image by 7.76. The experiment results show that the PSNR of the image restored by the proposed algorithm based on the PSFs combination has reached 28.09,which ensures the acceptable restoration quality for aerial multiple blurred images.

In order to improve the quality of the degraded images,an efficient local adaptive image denoising algorithm based on the Double-density Dual-tree Complex Wavelet Transform (DD-DT CWT) is proposed. The principles and characteristics of the DD-DT CWT are analyzed and a Bivariate Shrinkage Function(BSF) is derivated. Then,the noise image decomposition by the DD-DT CWT is implemented by applying four 2-D Double-density Discrete Wavelet Transform(DD DWT) in parallel and distinct filter sets in the rows and columns. According to the statistical properties of wavelet coefficients and the dependency of inter-level with intra-level coefficients,the BSF with local variance estimation is adopted to process wavelet coefficients and to reconstruct the denoised images by the shrunk wavelet coefficients. Finally,the proposed algorithm is tested on some gray and color noisy images. The experimental results indicate that,compared with the noise images,the Peak Signal-to-Noise Ratio (PSNR) gain of the proposed algorithm has reached 11.72 dB,Mean Structural Similarity (MSSIM) has been 2.7 times higher than that of noise images and the Composite Peak Signal-to-noise Ratio (CPSNR) reaches 11.68 dB when the noise variance is 30. Meanwhile,the algorithm is more efficient in noise removal and edge reservation for all the noise images with different noise variances,which improves the visual quality of the denoised images.

In order to solve the problem that the Very Large Scale Integration(VLSI) architecture of a Discrete Wavelet Transform(DWT) wastes a huge amount of hardware resources,a parallel VLSI architecture based on the DWT for JPEG2000 is proposed. The architecture introduces a (9,7) wavelet parallel filtering technique based on the time difference,so that the row processor and the column processor can process the signals in a parallel way. By using a 2×2 transforming module,several kinds of registers can be used to substitute a lot of medium transforming memories. Experimental results show that the proposed architecture can efficiently decrease the hardware complexity and can improve hardware utilization to 100% nearly under the tight critical path. The architecture has been implemented in a post-route VHDL,and can be used as a compact and independent IP core for Voyage Data Recorder(VDR) radar image acquisition cards.

The visual inspection based on an endoscope has played an important role in modern fault diagnosis of aeronautic engines. An inspection approach to surface cracks of the aperture on an aeroengine labyrinth disc is studied based on the industrial videoscope in this paper.After analysis on the appearance reasons of the cracks and their characteristics,an automatic recognition method for surface cracks is proposed based on a localization-recognition model and a novel fuzzy fast Hough transform is presented to detect the circular contour to locate the aperture. Then,an interesting region is defined around the aperture,and the image in the interesting region is binarized by a threshold and the binary image is thinned by a morphological method. Moreover,the chain codes in the image are analyzed to recognize the crack. The results of circular contour detection by the proposed fuzzy fast Hough Transform are compared with those by the Standard Hough transform and randomized hough Transform,respectively. The results show that the fuzzy fast Hough transform can detect the contour and can locate the aperture quickly and accurately. A crack detection experiment is carried out for 80 aperture images,and the results indicate that the detecting accuracy is 91.2%.

A correction method with high precision for Digital Camera(DC) lenses is proposed to meet the needs of the seamless mosaic of large scale images. The shortcomings of present algorithms are analyzed and a kind of detecting algorithm is proposed. In consideration of the structural characteristics of a chessboard pattern that the pixels in the images are central symmetrical for the points of intersection and antisymmetric for the lines of intersection of every two adjoining square panes,the precise coordinates of feature points can be decided by calculating the symmetry index and antisymmetry index for every pixel. Then,combined with the polynomial transform and bilinear interpolation techniques, a high-precision correction method for nonlinear aberration of DC images is put forward. The experiment shows the mean of horizontal error is 0.43 pixel and the mean of vertical error is 0.36 pixel after correction. These results show that proposed method has advantages in simple calculation,robust to noise,projective and distortion aberration and convenient for automatic detection. The correction and mosaic for a map photo is carried out,results prove further the effectiveness of the method by its application in the real map photo mosaic.

In order to suppress the pseudo-Gibbs phenomena around singularities of fused images and to reduce significant amounts of aliasing components located far away from desired supports when the original Contourlet is employed in the image fusion,a multifocus image fusion method in Sharp Frequency Localized Contourlet Transform (SFLCT) domain based on a sum-modified-Laplacian is proposed. The SFLCT,instead of the original Contourlet,is utilized as the multiscale transform to decompose the original multifocus images into subbands. Then,typical measurements for the multifocus image fusion in a spatial domain are introduced to the Contourlet domain and Sum-modified-Laplacian (SML),and the criterion to distinguish SFLCT coefficients from the clear parts or from blurry parts of images are employed in SFCLT subbands to select the SFLCT transform coefficients. Finally,the inverse SFLCT is used to reconstruct fused images. Moreover,a cycle spinning method is applied to compensate for the lack of translation invariance property and to suppress the pseudo-Gibbs phenomena of fused images. Using the proposed fusion method,experimental results demonstrate that the mutual information has improved by 5.87% and transferred edge information QAB/F has improved by 2.70% as compared with those of the cycle spinning wavelet method,and has improved by 1.77% and 1.29% as compared with those of the cycle spinning Contourlet method. Meanwhile,the proposed fusion method has advantages of good visual effect over the block-based spatial SML method and shift-invariant wavelet method.