A rapid detecting instrument of bacteria was presented based on the principle of Adenosine Triphosphate(ATP) bioluminescence.In the instrument, a high-precision A/D converter and a small Photomultiplier Tube (PMT) module were included to improve the optical detection sensitivity, and the optical, circuit and software designs were integrated to construct the rapid detecting system for bacteria combinated with homemade ATP bioluminescence biosensors.Standard bacterium suspensions were detected to study the relation of Relative Light Units (RLU) and the concentration of bacteria, and the Correlation Coefficient (R) is 0.976 0.In order to detect actual samples, an improved data processing algorithm was bring forward, and modified modulus were imported in the equation.Furthermore,the detecting concentration of bacteria in a hericium erinaceus sample was demonstrated by the proposed system, results show that the R is 0.993 compared with the result obtained by the standard Aerobic Plate Count (APC) method.The reproducibility of the system was tested and its Relative Average Deviation (R.A.D.) is 7.69 % and Coefficient of Variability (CV) is 9.07 %.The BCG vaccines were also detected by this system and the R is 0.997 compared with the result obtained by the APC method.its R.A.D.is 7.69 % and CV is 9.07 %.It is concluded that this system can complete the process of detection in few minutes, and can offer its high sensitivity and reproducibility.

It is a technical difficulty of fabricating laser resonator gratings that the -1 and 0 order diffraction efficiencies of the gratings are equal to expectation values synchronously in controlling zero plane width of grating groove by ruling technology.Based on the electromagnetic theory of gratings, the triangle model and design method of laser resonator gratings with 0 order coupling-out and -1 order oscillation are proposed according to the characteristics of TM polarization in non-anomaly region.The characteristics are that the change gradient of diffraction efficiency of the TM polarized wave in the region close to λ/d≈1.414 is small,and the diffraction efficiency is increased with a blazed angle and decreased with an apex angle monotonically.This model can realize the random energy distributions of 0 and -1 order diffraction efficiencies and can avoid the uncertainty of technique coming from old methods and reduce the fabricating difficulty.The model is used to fabricate a grating with diffraction efficiency of 65% and its efficiency error is only 0.6% at 10.6 μm.Moreover, the method can also be applied to design a grating with -1 order oscillation and -1 order coupling-out, so that the design method and fabrication technology of two kinds of gratings are unified.

A novel microwave/infrared dichroic beam combiner based on a high transparent metallic mesh was proposed to achieve the coupling of microwave and infrared signals in a dual-mode Hardware-in-the-Loop(HWIL) simulation system.Analysis models of microwave reflectivity and infrared transmissivity of the microwave/infrared dichroic beam combiner were then established.To verify the feasibility of the microwave/infrared dichroic beam combiner and the effectiveness of the analysis models, a sample of microwave/infrared dichroic beam combiner was fabricated by using laser directing writing technology on a Ge substrate,and the periods of metallic mesh under perpendicular and parallel polarizations were designed as 500 μm and 400 μm respectively to minish the difference of the microwave reflectivity at different polarizations.Moreover, the microwave reflectivity of the sample in 12-18 GHz was tested in a microwave anechoic chamber and the infrared transmissivity of the sample in 10-12 μm were measured by using a Fourier infrared spectrometer.The measurement and simulation results show that the microwave reflectivity of the sample is higher than -1.5 dB, and the infrared transmissivity is 83%.It can be concluded that the high transparent metallic mesh can be used as a novel microwave/infrared dichroic beam combiner in a dual-mode HWIL simulation system.

In order to achieve the athermalization of conformal optical systms, an infrared imaging system was presented by using the hybrid refractive/diffractive elements made of Si,Ge,and ZnSe and a diffraction surface mounted on the dome’s inner surface.The aberration characteristics of conformal optics and the aberration,thermal and dispersion characteristics of diffractive elements were introduced, then an athermal method by using the thermal compensation characteristics of the diffractive surface on the dome’s inner surface was given.A gimbaled conformal optical system used in mid-IR wavelength was designed.After analyzing the design,it is shown that the conformal optical system makes full use of the charateristics of diffraction elements in good phase compensation,small heat expansion coefficients and big dispersion factors and can work well at -40～70 ℃ while its average MTF value is higher than 0.43 in each Field of Regard (FOR).

The high-power CO2 laser disturbance to Long-wave Infrared (LWIR) HgCdTe detectors is investigated theoretically and experimentally in this paper.Based on the practical parameters of experiments, a theoretical model of temperature-rise is adopted to calculate the variation of temperature with the time and the power of laser irradiation on a detectors.Then, the high-power laser damage mechanism to the LWIR detector is discussed, and obtained results are compared with the experimental results from that a CO2 laser irradiates the detector on the distance of 15 km.The experimental results show that the power density of the CW CO2 laser with the power of 2.5 kW can achieve 0.16 W/cm2 after attenuation for 15 km distance.Therefore, the power density of the target after focusing by a lens can be 140 W/cm2 by calculation.Furthermore, when the power density of the target is 20.5 W/cm2, the detector can be jammed; when it achieves 110 W/cm2, the detector can be damaged and the detector temperature is high enough for Hg deposition.This experimental phenomenon well agrees with the results predicted by the theoretical calculation.The conclusions can provide references for the laser defense of detectors and laser jamming satellite-based detectors.

An all solid-state continuous-wave yellow laser at 570 nm is firstly reported.The yellow laser is from two Nd∶YAG spectral lines of 1 444 nm and 946 nm with a nonlinear sum frequency,and the two spectral lines correspond to their energy levels of crystal transition for 4F3/2-4I15/2 and 4F3/2-4I9/2, respectively.By using a doubly folded-cavity, the yellow laser at 570 nm is obtained by a intra-cavity sum frequency mixing of type-II critical phase matching KTP crystal at 1 444 nm and 946 nm.The outpower of yellow laser vs incident pump powers is measured,results indicate that when pump powers injected into two Nd∶YAG crystals are 24 W and 15 W, respectively, the 570 nm yellow laser of 560 mW can be obtained,and its power stability in 4 h is better than ±2.8%.Furthermore,the beam quality of the yellow laser is measured by a beam quality analyzer in an output power of 560 mW, and results show its beam quality factor M2 is 2.3 in the laser with a maximum sum-frequency output.

A laser interferometric method is proposed to detect the audible frequencies of underwater acoustic signals.A laser beam is used to irradiate the air-water interface when an underwater acoustic source causes the fluctuation of the water surface, the scattered beam with acoustic wave information will interfere with a reference beam according to the principle of the Michelson interferometer,and the frequencies of water surface waves can be concluded in the interference results.A precise optical detection device is used to detect the change of optical path difference caused by the water surface fluctuation and to achieve the real-time information concerning audible frequency of the underwater acoustic source by demodulating the data in processing system.It can be seen from the real-time frequency-domain graphs that the central frequency of the frequency band is the audible frequency of underwater acoustic signals, while the low-frequency band results from the Doppler shift caused by the natural fluctuation of the water surface.Experiment results prove that the method can be used to detect real-time underwater acoustic signals with the frequency from 4 kHz to 15 kHz and the test error is less than 7 Hz.These results can meet the requirement of the underwater acoustic signal detection for the real time and precision.

In order to improve the light transmission efficiency and to decrease the overlay of tricolor output spectra of a color filter, a 2D submicron embedded grating with a broadband width and high transmission efficiency was proposed.The device is consist of such four parts as a PMMA substrate, a ZnS layer, a metal submicron grating and a SiO2 layer.By the Rigorous Coupled-wave Theory (RCWT), the effects of the structural parameters on the spectral transmission characteristics were discussed, then three broadband color filters with its full width at half maximum (FWHM) about 100 nm were obtained by optimizing the structural parameters.For the red,green,blue color filters with the center wavelengths of 635 nm, 530 nm and 455 nm,their corresponding transmission efficiencies are 68%,78% and 71%， respectively.The numerical results show that the transmission efficiency has increased about 19% as compared to those of previous color filters, while the overlay of the tricolor output spectra is decreased effectively and the transmission characteristics are improved.

The principle of atmospheric sounding using a limb imaging spectrometer is analyzed.Then, a prototype of grating dispersion type UV/Visible limb imaging spectrometer is designed and manufactured for the spectral ranges of 540-800 nm (1st order) and 270-400 nm (2nd order).The prototype consists of a broad-band refractive achromatic fore telescopic optical system and a modified Czerny-Turner spectral imaging system and records images from 540-800 nm by the visible filter and 270-400 nm by the UV filter.The volume of the prototype is about 450 mm×250 mm×200 mm, and the total mass is about 8 kg.A spectral experiment is carried out with the prototype in a laboratory and the analyzed result demonstrates that the real spectral resolution of the prototype is 1.3 nm, which approaches to the theoretical spectral resolution of 1.12 nm and satisfies the technique requirement of 1.4 nm.The prototype has advantages in little mass and high precision and is suitable for spatial remote sensing.

A novel plasmon polariton nanolens formed by two subwavelength metal slits surrounded by surface dielectric gratings is proposed and demonstrated numerically.The dispersion characteristics of surface plasmon polaritons (SPPs) in a metal waveguide are studied by using genetic algorithm, which shows that the effective index can be adjusted effectively by changing the widths and dielectric constants of these slits, and the propagation of SPPs can be controlled further.By combining the Finite-difference Time-domain (FDTD) method with Perfect-matching Layer (PML) absorbing boundary conditions,the distribution of optical field in the lens structure is simulated and the effect of number of surface gratings on imaging characteristics is discussed.Experiments show that the focal length and width are increased with the increase of number of the surface gratings.When the number of the surface gratings increases from 5 to 11, the focal length increases from 1.715 μm to 2.325 μm and the focal width increases from 0.615 μm to 1.715 μm.Results show that this lens structure can be availabe for the future integrated optics.

The interrogation system based on a Fiber Bragg Grating sensor was designed for damage identification by using a unbalanced M-Z interferometer and a Phase-Generated-Carrier (PGC) technology according to the features of vibration type Structural Health Monitoring (SHM) method.The damage indicator,namely the sum of relative variation ratio, was constructed based on the wavelet packet node energy of vibration signals.Then,the control principle of statistical process was introduced and the index of damage identification was derived by using -R statistical process control.The 160 vibration signals of an aluminum simply supported beam under a health state and three kinds of damage states were measured using FBG interrogation,and results show that the duration of time domain vibration signal in each state is approximately 0.05 ms and their amplitudes are basically the same.The damage identification was analyzed based on the statistical process control limit (12.85, 41.35) ,and it is concluded that the damage identification system is good at continuously structural health monitoring.

In order to detect the oil pollutants in soils and to protect the soils, the fluorescence characteristics for crude oil,diesel oil and machine oil were measured by the fluorescence spectrum analysis method.Firstly, the stimulated emission mechanism of organic molecules in oil was studied theoretically, then the fluorescence spectrum experiments for three kinds of mineral oil in water and soils were conducted respectively by a fluorescent spectrograph, and the fluorescence characteristics of different concentrations of the mineral oil were investigated.It is shown from the measured results that three kinds of oil in the soils and water can all emit intensive fluorescence as they are excited by UV rays, and the fluorescence spectra are located between 350 nm and 700 nm.When their concentrations in water and soils are in the range of 0.0-1.0 mg/kg, the relationship between oil concentration and fluorescence intensity shows a good linearity and the minimum detecting limits are 0.005 mg/kg in water and 0.01 mg/kg in soils.These results demonstrate that the fluorescent spectral analysis can be used in the detection and analysis of the concerned oil pollutants in soils directly or indirectly.

A Zernike modal wavefront calculation method based on Computer Unified Device Architecture(CUDA) was presented for liquid crystal adaptive optical systems under the GPU general architecture.The wavefront calculation method was introduced and the CUDA characteristics were given.Then, a programming model for wavefront calculation by CUDA was established, in which it involved three kinds of optimized schemes including maximum threads, higher memory and transfer bandwidth.The method based on CUDA was tested and compared with the traditional method using CPU, and result shows that the consumed time by proposed method is less than 1 ms for a Zernike polynomial with 35 wavefromt values in resolution of 512×512, which means that GPU provides a computational power tens times greater than that of usual CPU-FPU combination.The method has reduced the system delay and improved correction speed and its optimized ideas for programming model can provide a reference for other computer models.

In order to compensate the hysteresis nonlinearity and to improve the precision of the nanometer positioning system with hysteresis in a piezo-ceramic actuator, this paper studies the inverse hysteresis modeling of the piezoceramic actuator.Taking both the wiping-out property and the modeling precision into consideration, a neural networks is proposed to realize the sorting & taxis model of hysteresis and to replace the reverse checking and interpolation method to reduce the error of the hysteresis modeling.A BP network with three layers is established,and the weight for every layer is obtained by training practical data.Based on the voltage and displacement extrema got from sorting and taxis, the input voltage of the piezoceramic actuator is obtained by using the neural network.Furthermore, several groups of experiment data are used to verify the accuracy of the proposed inverse model.Results indicate that this method using neural network reduces the average error of the input voltage to less than 1.5 V and the max error of the absolute value to less than 2.7 V.Compared with the reverse checking and interpolation method, this method effectively improves the precision of the Preisach inverse hysteresis model.

A novel absolute multi-pole magnetic shaft encoder structure is proposed to improve the resolution of multi-pole shaft encoders, realize absolute detection and to decrease their costs.A novel signal processing method based on calibration is used to eliminate the influences of nonlinear magnetic fields and assemble errors on the measuring precision.Two ring alnicos defined as an index track and a sub-division track are designed.The index track is magnetized according to a novel improved Gray Code,in which the linear Hall sensors are placed around the track evenly to obtain the absolute offset for magnetic region where the rotor belongs.The sub-division track is magnetized according to N-S-N-S, and the number of N-S is determined by the index track.Other three linear Hall sensors are used to translate the magnetic field to voltage signals,and the relative offset in a single N-S can be obtained by a look-up table.The magnetic encoder is calibrated using a higher resolution incremental optical encoder, the pulse output from the optical encoder and Hall signals from the magnetic encoder are sampled and transmitted to a computer at the same time,and the relation between them is obtained and stored in the FLASH of MCU by the look-up table.In the working state, the absolute angle can be obtained by the looking-up table according to Hall signals.A 12-pole magnetic encoder prototype is made and the result shows that its resolution can achieve ±0.72′and accuracy is ±1.2′.Furthermore,results also show that its accuracy can be improved by increasing the number of poles.

A double-layer micromixer made from Polydimethylsiloxanes(PDMS) was designed to realize uniform mixing for microlitre liquids.Its fabrication methods and the effects of geometrical sizes and Reynolds numbers(Re) on it were investigated.The design principles of the micromixer with main-assist channels were presented based on Fick’s first law and the velocity and concentration distribution of the micromixer in different geometrical sizes and Re were simulated by finite element method.The calculated results demonstrate that the mixing efficiency increases with the decreases of Re and width ratio and the increase of channel length.Then a mixer with a channel length of 9 mm, depth ratio of 0.71 and a width ratio of 1 was fabricated.The results of mixing experiments of water and red ink show that the micromixer designed at Re<5 can realize the fast mixture of two streams and the efficiency of mixture increases with the decrease of Re.It can meet the requirements of micro fluid’s rapid mixing at low Re.

A new calibration method of laser beam-direction was presented to implement any-orientation measurement for point laser sensors in reverse engineering.A calibration block whose calibration surface could be adjusted was designed to co-operate the calibration approach.During the calibration, the point laser sensor steped in an equal distance along X、Y、Z axes in the calibration surface, then the laser beam-direction could be confirmed according to the relationship between the stepping size and the length change of laser beam.Based on the CMM platform, the coordinate transformation process from the sensor coordinate system to the reference coordinate system was deduced when the point laser sensor was set in any-orientation, and the calibrating process was introduced in details.Finally, a comparison experiment on contact measurement and point laser measurement was carried out to verify the proposed method.Experimental results indicate that the error of the point laser sensor calibrated by the proposed method is (0.045 2±0.016 8) mm, which can satisfy the requirements of sampling accuracy in reverse engineering.

On the basis of research of L1B4 board-type horizontal ultrasonic linear motors, a kind of board-type horizontal ultrasonic linear micromotor in sizes of 36 mm×5 mm×4 mm is studied.The working principle and moving mechanism of this micromotor are described,and the miniaturizing design of the micro stator is executed by Finite Element Analysis(FEA) method.Then the vibration modes of the prototype are analyzed to realize the frequency degeneration.By using linear dynamic frequency response to simulate excited piezoelectricity, the layout of prototype with 5 PZTs and that of accustomed one with 2 PZTs are compared.Obtained results approve that the structure with 5 PZTs increases the vibrating displacement of an output point.The feature testing of prototype is carried out,and the experimental results show that this motor can offer the maximum velocity of 183.7 mm/s and the maximum thrust force of 50 mN, while excitation frequency is 62 kHz and voltage is 150 Vpp.This motor can be expected to be used in the linear driving field that has narrow space and needs certain driving force.

The mechanical compensation method is generally used to change the focal length of a zoom lens, in which the zoom cam is a key component to achieve the continuous zoom of the zoom lens.In order to improve the quality of the zoom lens and reduce the size of component，the priciple of changing focal length of zoom lens by the mechanical compensation method is researched and the structure of zoom cam is analyzed with the software of NX/Nastran.Then,the optimization design of the zoom cam structure is achieved based on the analysis results.The final optimization results show that the mass of zoom cam has tightened by 88 g, the wall thickness reduced by 1.5mm, and the size of outside diameter decreased by 3 mm.The modal analysis and thermal analysis are carried out on the optimized zoom cam,it indicates that the cam’s first-order natural frequency is 50.3 Hz by modal analysis and the deformation size of cam in the temperature difference of 90 ℃ is 0.006 mm by thermal analysis.Finally, through the lens transfer function test,vibration test and high and low temperature tests, the correctness of finite element analysis and optimization analysis are proved,which indicates that the optimized design of zoom cam is successful.

To overcome disadvantages of traditional grippers in more transmission chains, high energy consumption and complex structures, a flexible gripper using a type of smart material called Ionic Polymer-metal Composite(IPMC)was designed and manufactured.By using the Pro/E software to design the main structure of the IPMC gripper and to analyze its movement procedure, the whole IPMC gripper was assembled successfully.Furthermore, utilizing a type of laser displacement sensor, an electronic precision scale and a Canon digital camera, the tip displacement, tip force and average velocity of the four actuating IPMC membranes fabricated by ourselves for the gripper were tested, and then the data were sampled and analyzed by the Labview test system.Experimental results demonstrate that the turning angle of one IPMC actuating sample is over 180° within 3 V (when the voltage changes from -3 V to +3 V), the maximum tip displacement is approximately equal to its own length (except the fixed end),and this gripper can grasp an object whose weight is about 1.6 g (the mass of each sample with water is about 500 mg).These results show that the flexible gripper has advantages in simple structure, large deformation and low energy consumption, etc..Moreover, due to the flexibility of IPMC strips, the gripper do not damages the precision of the object’s surface because they can attach to the surface completely.Therefore, it is beneficial to grasping the object with high surface precision.

An atomic Force Photon Scanning Tunneling Microscope (AF/PSTM) is reported here,which uses a shift in resonance frequency to control the distance between a sample and a fiber probe.For self-oscillating tapping mode AF/PSTM applications, a sharp end and large conical angle optical fiber tip with an apex diameter smaller than 100 nm and a cone angle of 60～90° is fabricated.The tip is mounted on a piezo,then the piezo is placed in a positive feedback circuit to excite the probe to oscillate.The oscillation frequency of the tip is demodulated by a PLL, and its value is maintained a constant by controlling the negative feedback loop of the Z piezo to track the sample surface.Both theory and experiment are carried out to compare the externally excitation mode and self-oscillating mode.Results show that the self-oscillation mode reduces the response time of the tip in theory,and the bandwidth of self-oscillating mode in proposed AF/PSTM is 50 Hz in experiments, which is higher an order of magnitude than that of externally excitation mode response.A sample of grating is scanned by proposed instrument with a scanning rate of 1 Hz,and results demonstrate that the topography and optical images obtained by self-excitation mode have clearer contrast than those obtained by externally excitation mode.Besides faster response, the self-oscillating mode can improve the system resolution by increasing Q without effecting on the system response speed.

In order to research the influence of inner heat sources on the Mechanically Dithered Ring Laser Gyroscopes (MDRLGs), a temperature field model is established with finite element software ANSYS to simulate the temperature fields.Firstly, the finite element model is simplified, and then the disposal of material parameters, calculating methods of heat generation rate and the heat transfer coefficients are explained.Finally, the steady-state temperature field and the transient temperature field in 1 h and 3 h are given,which indicates the highest and lowest temperature regions and the distribution of a temperature gradient.A circuit to measure temperature using a platinum resistor is designed, and its precision is 0.005 ℃.The simulated results are compared with experimental ones and the error is less than 2%, which validates the rationality of the model.This research provides some helps for the selection of the temperature testing points, and improves the effect of temperature compensation for MDRLGs.

In order to reduce the influence of vehicular platform deformation on the theodolite measuring precision, compensate the angle error caused by large deformation, and to realize the trajectory measurement worked on mobile station, the principle of the effect of platform deformation on the measuring results was analyzed, and the center deformation angles caused by the platform deformation were measured by using dual-axis inclinometers put on the center of the azimuth axis, then the measuring coordinate deflection was calculated to figure out the deformation profile.The finite deformation testing results were storaged in a computer and a two-dimensional look-up table was created by taking azimuth and elevation angles as inputs.Finally, the any deformation angles on any azimuth and elevation angles were interpolated and calculated to complete the error compensation after the task.Results show that this method can compensate the error as large as 142″ caused by the platform deformation and can improve the azimuth precision by 44″ and elevation precision by 8.5″.It is shown that the method is a credible approach to pointing error correction of high precision vehicular platform theodolites.

A high-speed movement measuring machine controlled by a closed-loop system was developed to calibrate high-speed movement measuring devices.The machine could be set motion parameters and could do high-speed movement according to sine wave and triangular wave velocity curves.A granite service platform and aerostatic bearing sliders were used in the measuring machine and a linear motor and a PMAC motion controlling card were used in its high-speed drive system.Furthermore,the reflective steel grating scale was taken as a sensor to detect the displacement and feedback control.A rubidium atomic clock was used as the time basis of a high-speed data acquisition system,then the high-speed data acquisition of the displacement was implemented by the high-speed data acquisition system based on the Field Programmable Gate Array(FPGA) controlled by a high precision time interval generator.Experimental results show that the speeds of the measuring machine reach 5.3 m/s and 7.6 m/s, respectively, while it moves the distance of 300 mm according to the sine wave and triangular wave velocity curves, and the displacements following error range are -1.56-1.01 mm and -1.41-2.23 mm, respectively.These results show that the measuring machine can be used in the calibration of high-speed movement devices.

Tracking near-sea targets by the shipborne optical-electronic equipment may be failed for the swing of ship and sea-noise-wave.For precisely controlling the near-sea targets, a new algorithm is proposed for the automatic tracking of near-sea targets combined the CA model based Least Square Method (LSM) with the approximate-comparison model based on feed forward control technology.The algorithm firstly feeds the speed data come from an inertial angle to the speed loop by the CA model based LSM , then it is used to determine whether the sea-noise-wave is acquired or not by the approximate-comparison model.If the sea-noise-wave is acquired, the tracking mode should be switched from TV tracking to guiding tracking.An experiment is carried out, results demonstrate that the re-acquire time has been shortened to 437 ms from 5.39 m, which is higher an order of magnitude than that of original method.By proposed algorithm, an automatic tracking can be realized and the re-acquire time come from "dipping sea" is shortened greatly.

A testing system was established to control the liquid fuel temperature of Direct Methanol Fuel Cells(DMFCs).In view of the temperature controlling for a certain flux liquid fuel, a heating block was fabricated, in which a flexible stainless steel pipe was laid with a double-runway for heating up the methanol solution.The minimum length of the pipe and the heating power of the controlling system were calculated.Furthermore,semiconductor-refrigerating chips were utilized to heat and cool the liquid flow during testing procedures and a power actuation circuit and a transition circuit according to the controlling requirements of the semiconductor-refrigerating chips and the testing system were also developed.On the basis of PID based closed-loop controller, the liquid temperature was controlled automatically and accuratelly by a software module written in Visual C++.Experimental results show that the average cooling and heating velocity of the temperature controlling system is 14 ℃/min, and the precision of temperature controlling is better than ±2 ℃.It satisfies the system requirements of real time performance testing of DMFC.

A CCD camera calibration technology based on the translation of a Coordinate Measuring Machine (CMM) is proposed to decrease the calibrating cost of CCD cameras and to improve the calibrating accuracy, and its principle, mathematic model, calibrating steps and calibrating accuracy are studied.According to the CCD camera calibration principle and the high translation accuracy of CMM, the CMM brings the CCD camera to be calibrated to produce the relative translation with respect to the center of a white ceramic standard sphere (the calibration characteristic point)along the X, Y, Z axes,then the coordinates of the different positions of the calibration characteristic point in the probe coordinate system can be obtained.Furthermore,by capturing the image of the white ceramic standard sphere by the camera at every position, the coordinates of different positions of the calibration characteristic points in a computer frame coordinate system can be registered after image processing.By introducing the probe coordinate system, a calibration mathematic model is established, the calibration steps are given and a calibration system is set up.The comparing calibration result shows that precision of the CCD camera calibration system based on the translation of CMM is equivalent to that of the special calibration system , and the difference between the calibrating data of the two systems is within ±1 μm, which meets the requirements of the practical calibration for low costs,higher accuracy and higher calibration efficiency.

In order to improve the performance of Joint Transform Correlators(JTCs), the swarm intelligence method is used in a JTC to preprocess the joint plane image.Firstly, the concept of swarm intelligence method is introduced, and the feature of potential distribution and consistence with biological self-governing is illustrated.On the basis of above, an input plane image feature extraction method based on swarm intelligence is proposed,and then the basic theory of classicals JTC is introduced and its advantages and disadvantages are analyzed.Finally, the proposed image feature extraction method is used to preprocess the joint image in JTC.Processed results are compared with those of classical JTC and the JTC based on sobel operator in the extraction of joint image.Experimental results show that the crosscorrelation peak of swarm intelligence JTC is 4 240,which is larger 2 906 than that of the sobel JTC and larger 1 616 than that of the morphological JTC.These results show that preprocessing the input plane images by the proposed method is superior to those by traditional methods.

On the basis of analysis on vision systems for IC packaging, the micro-imaging systems, vision system calibration and pattern recognition locating algorithms are explored to realize the whole establishment of vision positioning system of IC packaging equipment.Aiming at characteristics of IC packaging, an experimental facility of vision system is established, and the system calibration is carried out.A series of pattern recognition locating programs are written with Matlab, and experiments are undertaken to analyze the precision and stability.Experimental results indicate that the calibration errors meet the requirement of normal distribution.The average errors of XY directions are 0.014 5 and 0.106 5 pixel, and the variances are 0.310 3 and 0.277 5 pixel.After system calibration and locating experiments by using a kind of IC chip, results show that the positioning error is less than 3.5 μm(0.4 pixel).The vision system has achieved a sub-pixel precision.

The output images from a conventional 4f information optical system are characterized by their limited frequency bands, random noises and coherent noises in coexistence.Therefore, when those images are restored,it is necessary to protect image details better and to remove the noises in high frequency domain and the coherent fringes in low frequency domain.According to these characteristics, a new method is proposed to restore 4f system images and to improve the image quality.Firstly, the step response image of the 4f system is used to obtain the prior knowledge of the coherent noise, then the step response and pending images are processed to remove the high frequency random noise in contourlet domain.Finally, based on the linear characteristics of 4f system, the pixel value of the pending image divides the pixel value of step response image from point to point to remove the coherent noise.Experiments demanstrate that the method can restore images and protect image details well and its Peak Signal to Noise Ratio(PSNR)and Structural Similarity Index(SSIM) have improved by 5 db and 6%, respectively.These results show that the proposed method is suitable for 4f optical systems.

A completely unsupervised defect detection method is proposed based on the Gabor filters and Mean Shift clustering to achieve the accurate automatic defect detection of a FPC gold surface.Firstly,the multi-dimension characteristics of an image to be detected are extracted by a series of processing steps including Gabor filter banks, morphological open and Gaussian smoothing.Then, the Principal Component Analysis (PCA) is used to reduce the pixel characteristics from multi-dimension to 2-D for reducing computation time in the next clustering.Finally, Mean Shift method is applied to cluster pixels with 2-D characteristics and the results can be divided into defect and non-defect groups to produce the binary image.The whole process needs to neither predefine the type of defects nor the texture type of FPC gold surface, which can be defined as a completely unsupervised method of detecting defects.A number of images of FPC gold surfaces with a variety of defects have been tested.Detection results show that the proposed method can accurately separate all types of defect regions from the background and has the characteristics of high stability and strong ability to identify defects.

A kinds of communication control board based on PCI bus is designed for improving the safety and stabilization of a photo-electrical detecting system.Firstly, the hardware design parts including the Field Programmable Gate Array(FPGA) circuit, PCI9054 design and other function modules are introduced and the PCI local bus control and the PCI system designed with the PCI9054 are studied.Then, other function modules used in the photo-electrical detecting system are shown.After finishing the design work, the communication control board is tested.Tested results indicate that the transmission velocity of PCI bus achieves 26.3 MByte/s, the least time unit is 1 ms with the accuracy of ±0.1 ms and the communication rate by eight series RS-422 is 230.4 kbit/s at the same time.The communication control board finished has been used in some important projects and has done well with high safety and stabilization under the complex environment.

In order to represent a scene exactly and entirely, an image fusion method based on Pulse Coupled Neural Network (PCNN) is proposed.After registering the images of multi-source sensors, obtained images are decomposed into several coefficients of low frequency and high frequency by using the 9/7 wavelet transform based on lifting scheme.The larger absolute gray values are selected to fuse low frequency images and the high frequency images are input to the PCNN, then a serial of fused sub-images can be obtained by comparing firing times after the iteration.Finally,the fused images are obtained by inversing transform using the 9/7 wavelet based on lifting scheme.By means of design of simulation experiments using visible and infrared images, the entropy, average gradient, standard deviation and space frequency are selected to evaluate the fused image.Obtained results show that the entropy, average gradient, standard deviation and space frequency of the fused image by using the novel fusion method base on PCNN are higher 0.010 4, 0.245 9, 0.113 1 and 0.284 6, respectively, than those by using the fusion method combining traditional wavelet and PCNN when a standard source image is used.

The precision of centroid location for star images is an important factor, which affects the correction of star identification and the precision of star attitude determination.In this paper, the error model using the traditional Center of Mass(COM) is analyzed and it is shown that there are systematic errors and random errors in the COM.For the systematic errors, this paper firstly analyzes the relation between the systematic errors and the calculated centroid position by a simulation, and then adopts the least square curve fitting algorithm to estimate parameters of bivariate polynomial and to compensate systematic errors.For random errors,it presents an edge threshoold method and a star image elimination method to remove errors.Furthermore,the CCD imaging model is analyzed and a Bivariate Polynomial Compensation Template(BPCT) is proposed.As the star intensity distribution is close to the Gaussian point spread function model and the BPCT is related to the dispersion size of star intensity distribution, the different BPCTs are calculated to compensate the systematic error of COM according to different sizes of star dispersion.The simulation results show that by using the BPCT, the accuracy of the star position can be enhanced from 1/20 pixel to 1/200 pixel,which meets the requrirements of star trackers for the high accuracy centroid location.

Illumination variation is one of the most significant factors affecting the performance of face recognition systems.As the adaptive smoothing estimation used in traditional Retinex algorithm can enhance the shadow edge falsely during smoothing face images, a novel Retinex algorithm based on adaptive smoothing with new a conduction function is proposed .This conduction function uses both spatial gradient and local inhomogeneity to measure the severity extent of pixel variation,which can smooth face images well without enhancing edge effects and lossing feature edges.During iterative process, the maximum between this iteration and last iteration is chosen to be the constraint,so this adaptive smoothing method with the proposed conduction function can be applied in Retinex theory to estimate illumination.Experimental results based on the Yale B face database show that the proposed algorithm can overcome shadows in side-illumination efficiently without losing feature edges of face images.The recognition rate is improved by 24.2% in the side-illumination condition at the best rate, and 4% in the no side-illumination condition compared with those of classical Retinex algorithms.The experimental results prove that the algorithm has illumination robustness,and can improve recognition rates under any illumination conditions effectively.