In order to solve the problem that traditional detection and tracking algorithms can not deal with head-on point targets effectively, an algorithm for head-on point targets detecting and tracking is presented based on infrared image sequences matching. Aiming at the characteristics of head-on point targets in the images sequences, the match result is marked directly in a point-to-point manner, the times of pixel points are counted which have been marked in the match process and an appropriate threshold is selected to show the real point targets. The algorithm merits lie in no need to suppose the doubtful target point in advance, and the tracking process does not change while the background complex degree or the number of point targets change. Especially, this algorithm can solve the problem of point target detection under the high noise environment. According to the simulations and experiments with real images, it has been confirmed that the method is feasible and valid.

In the optoelectronic system, the image quality is affected by the vibration, especially the angle displacement. In order to reduce the influence of the vibration on image quality, an irrotational displacement isolator is designed based on spatial mechanism and vibration theory, which mainly consists of base platform, moving platform, spatial linkage mechanism and vibration isolators. The mathematical model is formulated by the way of vector and coordinate transformation in the linkage mechanism so as to derive kinematics modeling of irrotational displacement isolator. Based on kinematics modeling, two models aiming at irrotational displacement isolator are simulated and analyzed. The result shows that the framework effectively isolates the vibration, restrains angle displacement and plays important role in improving the image quality.

Occlusion of the tracked object occurs frequently under complex background. At this time, a single track method can’t solve the occlusion problem. Under stationary background, object motion track can be used to calculate the future position when occlusion occurs. But there are motions of both the scene background and tracked foreground object on the moving platform, which results in complex object motion track. Aiming at the problem presented above, an object motion prediction method with background motion estimation is presented. The Kanade-Lucas-Tomasi (KLT) feature tracker is used to match corner features between adjacent pairs of video frames to obtain a sparse estimate of the optic flow field. A global, six-parameter affine motion model is fit to the observed displacement vectors to approximate the flow field induced by camera motion. Then, we use a Random Sample Consensus (RANSAC) procedure to robustly estimate affine parameters from the observed displacement vectors to compensate for the motion of the scene background. Finally, it is turned into object motion estimation under stationary background. Experimental results show that the new method effectively allows tracking through temporary, total occlusions.

Conventional Genetic Algorithms (CGAs) have the extensive application in image registration, but always have the problems of low accuracy and slow speed of convergence. By analyzing the mechanism of CGAs used in image registration, a notion called “retardation area” in GA search process was defined. Based on this point, we proposed a strategy of 8 Neighborhoods Search (8NS) in GA retardation area. This strategy forced the best individual to the direction of much higher fitness in GA retardation, which made the result of the image registration get higher accuracy. At the same time, the quality of gene was optimized. Hence, the quantity of similarity detection can be reduced and the speed of image registration can be speeded up. Compared with the exhaustive algorithm and CGAs, experimental results demonstrate that the 8NS-GA is faster and can achieve higher registration accuracy.

Since traditional tracking algorithms can hardly meet the requirements of both precision and low computational complexity simultaneously. It is hard to track targets with high speed and high maneuvering both in theory and practice. The Interactive Multi-model (IMM) algorithm with mixed filter between pattern space and measurement space was achieved according to the multi-rate tracking method, and the uniform representations of multi-rate Constant High-pass (CH) model and Constant High-High-pass (CH2) model were given. Meanwhile, according to the maneuver degree of the targets, adaptive "current" statistical model was introduced to describe the high maneuvering target, each model marched Kalman filter and particle filter respectively. And then, the tracking algorithm for high speed and high maneuvering targets based on multi-rate interactive multi-model algorithm was established. Experiment results show that, compared with traditional tracking algorithm, the method can efficiently improve the tracking precision with low computational complexity. It is expected to have great potential for engineering application.

A real-time correlation tracking algorithm, based on edge amplitude distribution, is presented to solve the aerial object tracking problem. We use the edge amplitude distribution histogram to express the target and track the target by computing the best correlation between the object model and the current object. The traditional mean shift algorithm requires a symmetric kernel, such as a circle or a rectangle, and assumes that the target’s scale is symmetric during the course of tracking. The pixels are used on the edge as the samples, about 5% to 10% of the total pixels in the kernel. The object model is updated by computing the Bhattacharyya coefficient. The experimental results show that the Bhattacharyya coefficient during the course of tracking keeps between 0.95 and 1.0, and algorithm runs well at 50 fps.

This paper put forward a method for analyzing effect of Earth Infrared (IR) radiation on exoatmospheric temperature measurement accuracy. Based on principle of 2-waveband IR temperature measurement, the calculation formula for color-temperature measured by IR sensors is derived and analysis method of measurement error is given when taking into account Earth IR radiation reflected by measurement targets. Then taking Ballistic Missile Defense (BMD) system EKV IR sensor’s temperature measurement on strategic ballistic targets as an example, the scale of effects on temperature measurement accuracy and discrimination capability of the sensor were analyzed when taking into account the Earth IR radiation by applying Monte Carlo method. Based on analysis above, it is concluded that the temperature measurement accuracy of the IR sensor on exoatmospheric objects is obviously degraded by Earth IR radiation.

The characteristics of infrared and visible signature of satellite depend on the situations of solar radiation in space, which is closely associated with the mode of motion and attitude determination of satellite. In the process of modeling optical signature of satellite in visible spectral band on first principles, optical signature is computed according to geometry relationship among satellite, sun, earth, sensor and optical properties of materials on satellite surfaces. Attitudes change of satellite is analyzed via mathematical description of sense of orientation to sun of solar vanes in proceeding orbit determination and ordinate transformation. The Sandford-Robertson model is then employed to calculate the Bidirectional Reflection Distribution Function (BRDF) which characterizes optical proprieties of materials on the surfaces of satellite, and formulae for calculating infrared and visible signature of satellite including direct and indirect reflections are subsequently deduced. Moreover, Elements of sun-synchronous orbit are used in simulation of optical signature of prototype satellite, and contours of radiation intensity are achieved as well as sequential gray scale pictures of radiance of satellite. Finally, characteristics of infrared and optical signature of satellite are analyzed.

To optimize the effect of conventional beam shaping algorithm for designing Phase Diffractive Optical Elements (PDOE), a new hybrid algorithm is presented. It combines the Genetic Algorithm (GA) with transformable scale BFGS (Broyden, Fletcher, Goldfarb, Shanno) algorithm, and penalty function is also used in the cost function conformation. The hybrid algorithm has the global nature of the genetic algorithm as well as the local improvement capabilities of the BFGS algorithm. Compared with other beam shaping algorithms, it has good efficiency, good convergence property and good stability. As a design example, both the conventional simulated annealing algorithm and hybrid algorithm are used into Gaussian beam shaping. To review the reliability of two algorithms, two indexes of diffractive efficiency and Signal-to-noise Ratio (SNR) are considered in numerical simulation. The design results show that hybrid algorithm has good convergence speed and cost function. It can acquire good diffractive efficiency and SNR with only a small quantity of iterative numbers. The shaping effect of diffractive phase optical elements designed by the hybrid algorithm is greatly improved. It will be widely used when the high uniformity is needed.

A method for measuring the thickness of photoresist, namely two-spectrum method, is proposed. Using AZ4620 photoresist throw at plane glass, the measuring thickness of photoresist is benchmark through ellipsometer. Compared with incident lights, the intensity change of emitted lights is examined by means of two spectrum method measuring. The purpose of measuring the thickness of photoresist is achieved. The result deviation is less than 2 percent. Compared with traditional methods for measuring the thickness of film, the new method has the advantages of simple computation and good maneuverability. As photoresist has the exposal characteristics, two-spectrum method is fit for measuring the thickness of photoresist.

Now, one of the main factors influencing the high resolution image is the dead zone. Firstly, based on the current static Volumetric 3D Imaging (V3DI) display systems, the imaging principles are described. And then, several reasons for dead zone existing are analyzed in detail, such as visual dead zone, voxel overlapping dead zone and voxel distortional dead zone et al. Meanwhile, in order to get a high resolution image, some effective approaches to solve these problems are given, such as structural design of integrated beam-splitter folded path for rear projection displays, imaging method based on rotating helix screen. Finally, a novel V3DI display system is built with the support of the theory of studying results. The experimental results show the system can weaken the dead zone effect, and acquire very high image quality.

By conjoining the wavelet coefficient absolute value with the local area energy, a new image fusion algorithm is proposed based on the techniques of wavelet decomposition. At the same selected location in wavelet domain of two source image, according to four kinds of configurations of the wavelet coefficient absolute value and the local area energy, the wavelet coefficients of fusion image are obtained respectively by selection and weighted average method. The experimental result indicates that the new algorithm is good in fusion result.

To improve spatial resolution of the image, a method for reconstructing high resolution image based on the independent equations of pixel grey level was proposed in this paper. An image sequence with low resolution was acquired through mechanical microscanning. The correspondence relationship of pixel grey level between the Image sequence with high resolution and the one with low resolution, which was acquired by displacing each time the image over the detector plane by a distance equal to half of the detector pitch, was analyzed. The independent equations of pixel grey level for high resolution images were set up, with which the high resolution image was reconstructed. Compared with other methods, results of simulation experiments show that the time taken in reconstruction with the algorithm is shortened from seconds to milliseconds. It lays the foundation for reconstructing super resolution video.

The parameter identification of the point spread function is very important for image restoration. Motion blur and defocus blur are the common blur types which degrade image. For the two blur types, a robust bispectrum-based method is proposed to estimate the blur parameters in noisy images. At first, we introduce the two blur types and traditional computational formulas. And then, the bispectrum of two blur types is deduced. In theory, bispectrum is not affected by noise, and corresponds to similar structure of blur function, so we can calculate the blur parameters easily by using the traditional computational formulas. The experimental results show that our method is effective in a certain range. In other words, when the blur parameters are not too large, our method can complete the parameter identification more accurately. When SNR is 25 dB，the deviation of blur parameters recognized is less than 1 pixel in this range.

To overcome the shortcomings of traditional distortion correction algorithms, an image distortion correction algorithm based on quadrilateral fractal approach controlling points is proposed. The standard raster image is used as the measurement target in the algorithm, the mathematical morphology is combined with sliding impending domain operation to fix on the distorted image’s pixel centroid, and the algorithm based on quadrilateral fractal approach controlling points is applied to fit high-order polynomial correction model. For image gray recovery, a two-step one-dimensional linear backward mapping method is used. The algorithm is applied on TMS320DM6437 DSP, and the experimental results show that for a 768 pixels×494 pixels image, the correction time is 0.036 s. The correction error is within 0.31 pixel and the edge information loss and inanition are effectively avoided.

In order to obtain the whole lens’ surface information, measurement of larger diameter optical plane was implemented by using smaller caliber interferometer. The problem of precise location was solved by the plane Givens transform of markers’ position. Each sub-aperture’s tilt and piston were removed by Zernike polynomial fitting. Global optimization stitching model was established. Nine sub-apertures stitching computer simulation was accomplished by using the model. The relative error values of PV and RMS are both about 10-6. A lens of 150 mm in diameter was used in nine sub-apertures stitching experiment. Compared with full caliber interference measurement results, the relative error value of PV is 0.36% and the relative error value of RMS is 2.27%. Simulation and experimental results indicate that the algorithm reduces the requirements of high precision rail in traditional sub-aperture stitching interference measurement.

In view of the single invalid data of the optoelectronic deflection meter system in the engineering practice of the bridge structural health monitoring system, which bring the illusive alarm and can not be identified by the traditional methods of signal processing and data analyzing，the paper analyses the characteristics of the single invalid data, and formalizes and describes them. And then, according to the features of the bridge structure changing slowly, a novel filtering algorithm is put forward based on data variety rate to identify the single invalid data, which can distinguish the single invalid data caused by sensors system from by bridge structure，and obtains better application effect for engineering practice.

In view of the high precision requirement on the intersection point and the angle of cross laser images used in industrial applications, a sub-pixel roof-type edge detection method is proposed for cross laser images. In a serviceable area, pixel-level roof-type edge points of bright strips are firstly detected based on maximum intensity in a local region. The sub-pixel edge point is then calculated by parabola fitting of the luminance of all pixel points near a pixel-level edge point in a certain range of horizontal or vertical direction. Finally, two lines of the cross laser image are fitted by least square method, and the intersection point and the angle of horizontal line of the laser image are calculated. For 50 samples (with a 640 pixels×480 pixels camera and 0.082 7 mm/pixel as resolution of the target) at distance of 4.5 m, the standard deviation is 0.013 4 mm for x, 0.015 8 mm for y and 1.950′ for angle, much better than that of pixel-level edge points. This method is used in straightness and twist-degree inspection system for rail guides and gets satisfactory results.

Based on principle of Digital Phase Generated Carrier (DPGC) modulation, laser interference vibration measurement system and its signal demodulation on software and hardware platform were established. Taking the 24 bit data acquisition device and 64 bit PC processor as the core of hardware circuit, real-time data acquisition of reference waveform generation and the signal for measurement was realized. DPGC algorithm was established by using LabVIEW software workbench, and then vibration signal real-time demodulation was realized. The experiment results indicate that the DPGC signal demodulation scheme based on PC can reduce drift and noise by traditional analog PGC hardware circuit such as analog multiplier and differentiator, enhance accuracy of signal demodulation and improve signal-to-noise of system. High accuracy measurement of the low-frequency vibration signals ranging from 10~200 Hz comes true. The measuring resolution is better than 0.14 nm and the dynamic range is 120 dB.

To get more representative iris features, three features, local feature point, local texture direction and changes between brightness and darkness of local texture are extracted through the gray-scale of iris images, which can depict the feature space of texture more fully and overcome the limitation of most previous algorithms which extract only susceptible single feature. Then patterns are categorized by the designed fuzzy inference regulation. The design of this piecewise linear classifier enhances the ability of linear classification of the algorithm. Experiments are implemented in two databases, respectively. The correct recognition rate is 99.41% and 99.67%, which demonstrate that many kinds of features can represent the variation details in the iris patterns properly. Therefore, the correctness is improved and the algorithm wins predominant recognition performance.

A facial expression recognition method is proposed. The Gabor features of facial expressional images are extracted and twice feature selection is used to recognize facial expression. Since the high-dimensional Gabor features are quite redundant, the high-dimensional features are sampled and the 2-D PCA is used to select valid features. Then, seven expressions are recognized based on multi-classifier combination with fuzzy integral. The experimental results on JAFFE facial expression database show that the proposed method is valid. It has higher computational speed and better recognition rate than 2-D PCA.

A design method for high-resolution and high-speed digital video camera, which supported 2.0Mega Pixels and 15 frames per second, was presented. How to solve some problems of key technology on the video camera was introduced. ICX274 was used as CCD sensor, AD9923A was used as the analog front end and other chips included XC3S1200E and TE3310RPF. The AD9923A converted CCD analog signal to digital signal. The digital signal was sent to XC3S1200E which converted Bayer signals to YUV signals. The YUV signals were inputted to TE3310RPF to generate JPEG data. AT91RM9200 sent the JPEG data to network. Many key technologies which included CCD timing control, Bayer signal to YUV signal conversion, TE3310RPF control, and data transmission in network were involved in this video camera design. How to solve these problems were also introduced. The testing results show that the design method can collect, compress and transmit 15 frame 2.0Mega pixels image per second.

Since it is difficult to measure the complex surfaces of large scale products precisely, the market had been occupied by foreign commercial measuring software for long times. The close range industry photogrammetry system, namely XJTUDP, has been developed successfully at home firstly. Composing and explored contents of this system were introduced. The high accuracy center measuring method of marked points technique based on the fitting sub-pixel edge, and the methods of coded points designing and coded points auto-detecting were investigated.The camera calibration techniques based on photogrammetry and 3D reconstruction techniques were studied emphatically, including co-linear equation, image orientation based on coplanarity equation, direct linear transformation solution, the outer polar geometric constraints and bundle adjustment solution. Both XJTUDP system and TRITOP system were tested individually by taking VDI/VDE2634 testing program as reference project and taking self-designed framework of cube as tested target. It is concluded that their precision was less than 0.1 mm/m, which proves that photogrammetry system explored by ourself is applied in field of industrial measuring successfully.

According to the characters and requirements of the nondestructive steel radial tire detection, an X-ray inspection and defect recognition system for all-steel radial tire is designed. The sector X-ray and linear CMOS camera are used to capture the non-blind area and full field radial tire x-ray digital image, and the lateral resolution of the image is 0.08 mm. Based on the character of the radial tire, the defect recognition algorithms are designed to segment and detect the inner defect of the radial tire automatically in the meantime of recording. Experimental results show that the defects of the tire with dimension ranges from 23 to 35 inch, such as bended Cord, crossed cord, wide spacing cord, broken cord, belt defects, impurity defects and so on, can be recognized on-line by the system.

The Forward Passage Gain (FPG) is an important factor that affects the detection Signal-to-noise Ratio (SNR) and dead band nonlinearity of Fiber Optic Gyroscope (FOG). If FPG's value becomes lower, the SNR and dead band nonlinearity will be worse. The mathematical formula of FPG is deduced from the physical model, and the influence of space environment on the FPG is analyzed. The pseudo-random identification method is used to on-line estimate the FPG of the FOG. The pseudo-random signal is put on the sawtooth as the identification input signal. The pseudo-random demodulation works to extract the FPG. At the same time, the square-wave demodulation works to extract FOG's error signal. The correlation among modulated error signal, square-wave demodulation signal, pseudo-random disturbance and the pseudo-random demodulation signal is calculated, and the conclusion is made that pseudo-random demodulation will not bother the working of the FOG's demodulation process. The hardware realization is introduced and during the radiation test of the SFS components, the pseudo-random identification is validated. The test result shows that FPG can reveal the changing of output power from the SFS and the Random Walk Coefficient (RWC) of FOG.

During measuring the thickness of metallic foil with tandem differential white light interferometry, a new data processing method for white light interference spectrum based on the Minimum Mean-square Error (MMSE) was proposed to overcome the poor anti-noise capability of traditional extremum method. In the new method, the spectrum measured value was firstly normalized to eliminate the intensity modulation, and an initial value of Optical Path Difference (OPD) was calculated by extremum method. The mean square deviation between the normalized measured spectrum and the theoretical spectrum from the initial value was calculated. When the estimated value of OPD is closest to the actual one, the mean square deviation will be minimum. The optimal estimated value was obtained by means of iteration. Emulation and experiments show that the data processing method has higher precision and better Signal-to-noise Ratio (SNR) than the traditional extremum method.

By analyzing the methods of imaging spectrometer image data compression, the key point of interference figure compression is found out, and the corresponding relationship between interferogram matching and spectrogram matching is analyzed using the basic principle of several spectral matching models. Then, we present a method to determine the maximum compression ratio of interferogram by using the relationship between interferogram matching angle and spectrogram matching angle. At last, the reliability of the theoretical and technical basis of the method is validated through related experiments based on real spectrum data and its corresponding interferogram, which offers a solid theoretical support for the high efficient and robust modeling of spectrum data compression, and supplies a new idea for the development of interferogram compression method with high compression ratio.