To reduce influence of quadrant detector output signal nonlinear error on the system precision, a matched condition was given to ensure maximum linear interval of detector output signal, and characteristic optimization of detector output signal by subsection linearization method was put forward. In the paper, a mathematical model of quadrant detector output signal was established by light energy integral. Based on the model, the relations were analyzed between linear characteristic of signal with radius ratio of spot to photosensitive area, signal sampling precision and non-uniform of photoelectric transform coefficient and so on. Then, matched condition was acquired which is most proper for radius ratio of spot to photosensitive area corresponding to maximum linear interval of detector’ output signal. And the optimum linearity characteristic parameters of detector’ signal were gotten by subsection linearization fitting. The simulation result shows that, compared with multi-calibration and least square fitting in entire linear interval methods, max and mean of nonlinear errors in entire linear interval are nearly reduced to 1/10 of the method’s mentioned previously, and accordingly, the linear interval of detector is expanded 20%~30% under the same condition.

A 3D measurement system of complex deep-hole profiles based on ring-structured-light was presented. By analyzing disadvantages of the original system, such as nonlinearity between the input and output, the exterior and the interior stripe points of the image belonging to different measured profile in once measurement and existing occlusions due to rifling, an improved scheme was proposed, in which the half-angle of the conic mirror was changed from 45o to 50.7o. In the meantime, related parameters of the measurement system were also adjusted. The improved system eliminates nonlinearity error, which increases measurement accuracy. The exterior and the interior stripe points belong to the same measured profile, and one side of the occlusions of the exterior stripe is eliminated, which can increase the efficiency and the accuracy of 3D reconstruction.

In order to measure the radial distortion in optical imaging system, a new method based on carrier-fringe pattern phase analysis using dilating Gabor transform was presented, and measuring radial distortion was converted into measuring modulated phase. Firstly, a deformed grating pattern (namely a distorted image) was obtained from a vertical original grating pattern by the optical imaging system. Then, the dilating Gabor transform was used to extract the fundamental frequency and phase at the central point of the deformed pattern, where there was no distortion, and the fundamental frequency information was the fundamental frequency component of an ideal image. The deformed grating pattern was analyzed by use of the dilating Gabor transform, and the modulated phase distribution, which was used to calculate the radial distortion distribution, was obtained. Finally, by use of the above radial distortion distribution and bilinearity interpolation, the corrected image was reconstructed. Detailed theoretical analysis is presented and experimental results demonstrate the validity of the above method.

In order to test 3D profile of micro/nano structures with high resolution, a measurement system was developed based on microscopic interferometry and polarization technique. First, five interferograms with π/2-phase increase were acquired by five-step phase-shift interferometry. Then wrapped phase maps were calculated with Hariharan five-step-phase-shift algorithm. Finally, 3D profile of micro/nano structure could be gotten with branch-cut unwrapping algorithm and the phase-height formula. A polarizer provided a polarization beam and adjusted its illumination power. The polarized beam was parceled into two perpendicular directions by a polarization-beam splitter, and the intensities proportion of the two beams could be adjusted by rotating a 1/2-waveplate making it possible to compensate the differences of the reflectances between the sample and the reference mirror. This arrangement allowed that the interferograms had the advantages of high contrast and optimum intensity, and this was useful to increase system resolution. The system’s main performance parameters include the Ra-repeatability-measurement resolution better than 0.06 nm, the error indication better than ±1%, and the variation of indication better than 0.5%. The experimental work conducted on several typical micro/nano structures such as standard multi-indents structures, Si-microphone membranes and Si-micro-gyroscope folded-beams confirms that such a measurement system is efficient to characterize 3D profile of micro/nano structures with high resolution.

An optoelectronic method of high-accuracy small-angle rotary measurement was introduced based on multiple-reflection between two plane mirrors, and an integrated experiment system was constructed. Then, the relationships between the system’s measure precision and measure range with laser beam reflection times in plane mirrors was studied. Finally, an actual application example was calculated and analyzed. The analyzed result shows that the measurement method has the characteristics of high precision and simple structure. When the reflection times are above 10, system measurement precision will reach 0.01 arc second. It provides a new method for sub-second accuracy level dynamic angle measurement.

An analytical design and parameters tuning approach of Internal Model Control-PID (IMC-PID) controller for an opto-electronic tracking system with time-delay is proposed in this paper. Firstly, a first-order delayed integrating (FODI) model for the system is built, and the model is approximated by a Second Order Plus Delay Time (SOPDT) mode. Then, the time-delay term in the model is replaced by the simple first-order Taylor expansion. So, the tuning rules of the controller parameters are founded. Especially, the single adjustable parameter λn of the IMC-PID controller can be calculated by choosing an appropriate maximum sensitivity to guarantee robustness of the system. The simulation results show that the proposed method could provide a better dynamic performance of both the command tracking and disturbance rejection, and have better robustness against parameters perturbation than conventional method. In addition, the experimental result demonstrates that the method could bring better tracking performance and higher accuracy than the PID control.

To overcome the problem that the common subtractive clustering object locating method could not acquire video object’s scale and orientation parameters, an improved subtractive clustering object locating algorithm, that is called the scale and direction adaptive locating of video moving objects with subtractive clustering algorithm, is proposed. The proposed method first uses common subtractive clustering locating method to acquire video object position and moving object number of each frame. Then, the proposed algorithm uses fuzzy C-means clustering algorithm to further cluster the moving object foreground pixel samples. Finally, matrix analysis theory for calculating eigenvalues and eigenvectors of covariance matrix is applied to compute the object scale and orientation parameters. Experiment results show that the proposed object locating algorithm could obtain much more reasonable video object locating results.

Gyro’s three key parameters, output noise, resolution and phase response are calibrated in an aerial camera by using two types of devices which are commonly used in engineering. First, according to datasheet, two gyros’ noise power ratio is 784:1. By adopting a new scheme suiting for practical application, calibrated noise power ratio is 74:1 owing to system noise. Second, relationship between gyro’s resolution and analog-to-digital converter is presented. Thirdly, since there is seldom discussion about phase response in gyro datasheet, a flexible and novel experiment platform is designed, and then phases response in 0~16Hz is calibrated with error less than 5%. Finally, effects of gyro parameters on servo control system are thoroughly analyzed based on above results. Experiments indicate that the gyro parameters calibrated precisely in actual equipment improves system performance and stabilization greatly.

An efficient approach based on wavelet transform is presented to detect the variation degree of dynamic background. Firstly, some candidate regions can be obtained by processing IR data through wavelet transform, then the degree of the difference between current and referenced background can be judged through detecting the difference of the above feature regions. Finally, some small moving candidate targets can be detected by eliminating background. After eliminating background, data association and robust tracking of those targets can be realized by level set tracking method based on region-based active contour with dynamic priors. Dynamic priors include shape descriptor, intensity information, moving features and etc. In addition, targets’ merging and splitting phenomena occasionally arise during multi-target tracking. This problem can be solved by ‘memorizing and filling’. The proposed approach is validated to track multi-target effectively by using actual infrared image sequences with complex dynamic background. Experiment results indicate the feasibility and effectiveness of the proposed method.

To enhance Camshift algorithm on the adaptability of high-speed object, complex background and covering, an improved algorithm combining inter-frame difference and motion prediction is proposed. First, before calculating back-projection image, the color information is associated with the motion information gained from inter-frame difference to eliminate color noise and automatically initialize tracking. Second, in order to make the algorithm applicable with high-speed object, position of target is predicted and searching window is updated according to tracking state. Experimental results prove that the improved algorithm is robust to high-speed target, covering and color noises.

The mechanism of precision tracking turntable was designed. The entity model of the precision opto-electronic tracking turntable frame was established and the static and dynamic analysis of the precision turntable was done by using the software ANSYS. The static analysis results show that the intensity and rigid requirements of tracking turntable frame were satisfied. The mode analysis indicated that the precision opto-electronic tracking turntable frame had good dynamic characteristics, and syntony did not occur. The rationality of turntable frame was validated by analysis results, and it provided the important evidence for the optimization design of precision opto-electronic tracking turntable.

For the multi-target tracking problem of complex backgrounds, the models of appearance, disappearance and occlusion of target in observation scene were described, and a probabilistic multi-target tracking algorithm based on particle filter was proposed. The variable number of targets was modeled by a Markov chain, and a hidden variable was augmented in state representation to represent possible occlusion explicitly. An exclusion principle based on observation likelihood was constructed with the spatial histogram of the target, and the state and number of targets were estimated by the posterior probability. The experimental results show that the proposed algorithm is robust to the problems, such as variable target number, the similar appearance disturbance, and short-time occlusion.

According to the Randomized Circle Detection (RCD) algorithm, the circle parameters needed to be calculated four times for each four points sampled randomly. A fast randomized algorithm for detecting circles was presented. If the distance between one point and the boundary of the circle determined by the other three points was bigger than a certain threshold in the process of judging whether the four points could determine a candidate circle, the randomized sampling and computation should be carried on again. The proposed algorithm largely decreased the calculation times of circle parameters by the new threshold. It affirmed the candidate circle for true circle to apply a quick evidence-collecting process. Some synthetic images with different levels of noises and real images were taken to test the performance. Experimental results demonstrate that the proposed algorithm can detect circles more quickly than RCD, and has a strong robustness.

Automatic face recognition is a challenging problem in the biometrics area, where the small sample size problem exists. An Enhanced Relation Discriminant Analysis (ERDA) method is proposed to solve the small sample size problem. In our framework, the neighbor and class relations of data are used to construct the embedding for classification problems. The proposed algorithm learns the embedding for the submanifold of each class by solving an optimization problem. After being embedded into a low-dimensional subspace, data points tend to move due to local intra-class attraction or inter-class repulsion. ERDA aims to map the image space into a submanifold that faithfully discovers the local discriminative manifold structure of face image. This method accounts for both the representation and the classification points of views. Experimental results on the AT&T and Yale face image databases demonstrate the effectiveness of the method.

Design and realization of tea component analyzer is presented based on Near-infrared (NIR) spectrometry. The discrete interference filters are used as NIR diffuse reflectance part. A white board is embedded into the rotatable sample cell. The samples and the white board are scanned in a fast and alternative way, which tests sample and the white board simultaneously, effectively simplifying the mechanism and accelerating the sample detection as well as improving the accuracy. InGaAs Detector, instrumental amplifiers and 24-bit A/D converters are used in photoelectric signal processing system. The computer completes the value determination of absorbency and tea component analysis with the data transferred from USB 2.0 interface. The relative error between this instrument and conventional chemical analysis is less than 5% by the experiments result.

A novel distributed sensor system based on Fiber Bragg Grating (FBG) sensing technology is proposed. In huge project, there are often one hundred positions to be monitored. If these positions are divided into several groups and every group has four fiber Bragg gratings with the same wavelength, it can implement one hundred fiber Bragg gratings on one fiber. If the temperature of one point is varied, the largest temperature of the group can be obtained. The system will alarm if this temperature is over set value. Based on the theory analysis, temperature experiments of different wavelength have been done. In 10～110°C, the central wavelength of FBG varying with the temperature has fine linearity. It has been seen in experiments that the system has the characteristics of the high measured precisions, good reliability and low cost in distributed multimetering.

To meet the requirements of positioning accuracy of tridimensional mapping system, based on development of three-linear CCD camera, the practical imaging model of three-linear tridimensional mapping camera was constructed to confirm mathematics formula between the ground point in earth coordinate system and its image point in CCD sensor coordinate system and in uniform image plane coordinate system. Then, the practical error model was deduced to analyze main error source, and integrated positioning accuracy was analyzed according to camera parameter. The results indicated that the positioning accuracy can reach 50 meters under the condition of non-control ground point, meet the requirements of user and verify validity of imaging and error model.

The central wavelength of the polarization modes will separate obviously when the multiple cavities narrowband thin-film filter is in oblique incidence, which will cause the rapid increasing of the polarization dependent loss. It will seriously affect the optical communication system performance. The reason of the polarization dependent loss in oblique incidence is analyzed theoretically. By optimizing the thin-film stack, the central wavelength of the two polarization modes can be calibrated and the polarization dependent loss in the passband of the filter can be restrained effectively. Meanwhile, by using the polarization beam-splitters, the polarization dependent loss in the whole bandwidth of the filter can be also restrained. The experimental results show that these two methods can restrain the polarization dependent loss of the narrowband filter within 0.2 dB in oblique incidence, and they can be used in different occasion according to different practical demand.

The micro-channel heat sink is an effective way to solve the heat removal problem in high power semiconductor laser array. The temperature in a high power semiconductor laser is analyzed by the finite element method. The temperature is obtained in a single and 3, 5, 9 micro-channels heat sinks with the space of 100 μm and transverse size 200μm×60μm. The effect of number of micro-channels on maximum temperature is presented. The result shows that the temperature in the semiconductor lasers in 36 A injected current stable operation can be cooled down to 306 K by the heat sink made up of 9 micro-channels, while the temperature is dropped to 342 K for a single micro-channel. The temperature distribution is also simulated with increasing the space between micro-channels. It is found that the temperature can be dropped to 308 K by 5 micro-channels with the space of 260 μm.

In order to improve the frequency stability of laser in 1.5μm regions, two acousto-optic modulators were used to shift the laser frequency and the absorption spectra of an acetylene transition (12C2H2 P9) at 1530.37nm was used as the frequency reference. Then, the dispersion-like frequency discriminating curve was obtained by subtracting the two shifted absorption spectra of acetylene. The new method could avoid the extra frequency noise due to indirect frequency modulation of laser source in a conventional scheme. After locking laser, the typical frequency jitter was less than 5.8MHz in 50s, which is significantly better than the frequency fluctuation under free-running condition. Hence, the frequency stabilization of external cavity diode laser with a modulation-free output can be realized.

Since the Deuterium Fluoride (DF) chemical laser utilizes a high-speed reacting flow with multiple chemical species, the choices of key species, reactions and rate constants have a great impact on the simulation result. Different types of DF chemical kinetic model are discussed, the simplified rules are given, and several simplified kinetic models are used in the numerical simulation of DF chemical laser. Through comparison between these results and the results under base kinetic model, the differences in the spatial distribution of small signal gains and DF (v) (v=0, 1, 2) are obtained. Finally, aimed at (NF3+H2+He)/D2 and “cold reaction” DF chemical laser system, an 11 species and 23 reaction simplified kinetic model is recommended, which can decrease computational costs and make the programming of reaction flowfield simulation more easier.

Time-varying velocity measurement by self-mixing laser diode is studied. A signal processing based on polynomial phase parameter estimation is presented. Based on the principle of self-mixing laser diode velocity measurement,polynomial phase expression of the laser diode output is deduced when the self-mixing laser diode is used to measure time-varying velocity. Namely,a model is proposed for the simulation of time-varying velocity measurement. At last,the signal processing based on polynomial phase is used to obtain target’s velocity and acceleration . The simulation results prove the validity of the algorithm even in the situation of low SNR.

Disadvantages of wavelet, ridgelet, curvelet and other methods for the restoration of line similar pixels in noisy image are analyzed. Then a structure and standard uniformity of directional operators for anisotropy detecting are presented. By using coarse DWA scheme, line similar line pixels in noisy image are restored through the schedule of detection, wash out and weight average. With the program of shrinking, connection and recording, line similar line pixel clusters in noisy image are segmented from each other by means of SCR method. The rationality of directional operators, accuracy, effectiveness, limitation and consistency of relative restoration and segmentation methods are validated by experimental simulation.

Median filtering can get a good real-time background image from a short time video-steam if the moving objects are not very crowded. By using median filtering method, the background real-time update of the moving object detecting was studied based on video information, and the selection of samples of image series was researched. Moreover, the theoretical model was established, and a good background image was got. The results show that, if the time of image series samples is too short, the background image will have objects shadow. If the time of image series samples is too long, the background image is not update, which will cause errors in the detecting of the moving objects. The time length problem can be solved by selecting the appropriate image series samples based on the theoretical model proposed.

Noise in fundus images seriously affects the diagnosing of lesions. On the basis of prior knowledge about fundus images, four multiscale geometric transforms are adopted for de-nosing, which are respectively Ridgelet, Ridgelet combined with Wiener filer, Wavelet,Contourlet. The Local Mean and Local Standard Deviation algorithm is given based on Gaussian wave extraction, which is used to estimate Signal-to-Noise Ratio(SNR) of the processed images and give an objective quantitative evaluation to treatment effect of the above de-noising algorithms. The results show that the image processed by Ridgelet combined with Wiener filer is the most clearest, and its SNR improves the most obviously, about 5.04 times compared with the original image. The results of objective quantitative evaluation are in accordance with subjective visual feeling.

Aiming at online impurity detection of ampoule injection, a system based on machine vision is developed and analyzed. The hardware structures, configuration of imaging card and so on are discussed. According to the requirements of application, we designed a reasonable imaging method for the detection. Moreover, the way of high spinning-emergency stop was adopted to successfully distinguish the impurity from the blemish. With regard to the software, an oriental fast and effective way of segmentation is proposed as to the low level gray of the impurity and an effective way to detect impurity based on machine vision is achieved. At last, the experiment result was given, which proves the feasibility of the program.

For the accurate registration of high resolution Synthetic Aperture Radar (SAR) images, local transformation functions are preferable to global ones. In this paper, a local mapping model, namely neighborhood reconstruction model, was presented to fit the geometric distortions in SAR image registration. First, each point in the reference image was characterized by the neighboring Control Points (CP) with reconstruction weights. Then, a criterion for neighboring CP selection was proposed to minimize registration error at each point. Finally, the associated point in the sensed image was resampled according to the geometric distortions and the reconstruction weights. The theoretical support from neighborhood reconstruction model to the classical piecewise linear approach was also presented. Experiments on both simulation data and real high resolution SAR images show that the registration accuracy is improved.