A multi-feature vision tracking method by CamShift optimizing particle filter was proposed. The position and scale of the particle were optimized by CamShift, and the algorithm can select the proper size of the tracking window in the scenarios that the object scale varies. Meanwhile, by adopting an adaptive method, the color information was combined with motion information in the framework of CamShift optimizing particle filter. Because the propagation of the particle was optimized, which could make each particle converge on the true state region of target, the effectiveness of particle is improved. The experimental results show that the tracking error of CamShift optimizing particle filter with only 10 particles is less than that of the traditional particle filter with 100 particles. What's more, as a result of the use of multi-feature, it can still track the target accurately and solve the problems encountered in tracking an object with illumination variation and the background color clutter. Stable tracking can be realized by a few particles, which can reduce calculation cost and improve the robustness of tracking.

In order to achieve opto-electronic system precision pointing, a structure of horizontal opto-electronic system, which was fixed on soft-rope hanged platform with attitude stabilizing device, was presented. According to the kinematics relationship of the parts of this system, some methods were employed to describe the topological configuration of the system, such as matrix of channel vector and matrix of constraint force element. Friction torque, platform attitude control torque and other disturbance torques were discussed. Lagrange method was used to establish the multi-body system dynamics model. Simulations were performed to study the influence caused by random rotation of the platform in the azimuth axis and pendulum in the plumb-horizon plane to the Line of Sight (LOS) pointing accuracies. It can be concluded that platform attitude controller is unnecessary when pointing controller is strong and the anti-twist system works well. Furthermore, pendulum would lead to coupling of azimuth angle and horizontal angle of LOS.

In the localization process of multiple passive sensors, constraint formulas of trigonometric functions are given by means of the relationship about sensors. Measurement errors can be obtained by optimization method. Most cross points do not overlap because the results are the local optimal or approximation optimal. Not only the sensor’s performance but also the radial range influences the target localization error, so the two factors are considered in this paper. The sensor’s performance and radial range are regarded as parameters. Each cross point weight is calculated, weighted sum of cross points is given, and then better localization results are obtained. Simulation results show the modified algorithm is efficient.

While applying Sliding Mode Control (SMC) algorithm, the time and space delay and dead zone may cause the unexpected output vibration in actual controller design. In this paper, the damage to actual system and the causes of its existence are analyzed. Considering the inherent random disturbance and noise in servo system, a SMC algorithm with exponent approach law based on Kalman filtering is proposed. The simulation shows that compared with common sliding mode algorithm, the new algorithm can not only suppress the vibration but improve the dynamic quality of servo system.

Different from the conventional spherical compound eye imaging system, a novel optical design structure was proposed that the non-homogeneous microlens array was fabricated on the spherical substrate. The whole array was arranged as symmetrical circle pattern, and the focal length of microlenses arranged along the radius direction was decided by its position and equal to the distance between substrate and photo-detector. According to the imaging theory of geometrical optics, the design parameters of microlens array were computed and verified by ray tracing procedure. From the simulated results, we can get the conclusion that the optimal design structure of single-layer spherical compound eye can focus the whole field of vision on the photo-detector and improve the imaging quality of the edge field of vision obviously. Moreover, the fabrication technology of non-homogeneous microlens array on spherical substrate is discussed.

Dill’s classical exposure model on depth axis and time axis are generalized, respectively. On the depth axis, based on the Kirchhoff diffraction equation, complex-number-refraction-index was introduced, and the diffractive illumination distribution inside the resist at one moment was calculated by using the Beam Propagation Method(BPM). On the time axis, the characteristic of SU8 and its reaction course during exposure were analyzed, an appropriate chemical reaction kinetics model of the photocrosslinking reaction was built, and the diffractive illumination distribution at different moment was calculated. Finally, by building a kind of integrated exposure model, the diffractive illumination distribution after some exposure time was calculated. The results indicate that the illumination distribution variation with time inside the resist is not obvious and the impact of exposure time on illumination distribution is relatively low, which will be expanded during the post-exposure-bake process.

A Langmuir-Blodgett (LB) inducing method was firstly used to prepare single layer and multilayer conducting composite PEDOT-PSS film. The application of PEDOT nanostructure to improve organic electric device performance of Organic Light Emitting Diode (OLED) was studied. Furthermore, these conducting polymeric LB films were taken as the hole transfer layer in OLED device, and an efficiency enhancement of carrier injection was observed, which was ascribed to the ordered structure of these film. However, a further investigation shows that this PEDOT-PSS film has inferior structure ability.

According to the theories of thin film calculation and optimization, a new global optimization method for multilayer film thickness calculation is presented. Firstly, a new group of local optimums is found out through genetic and mutation operation. Then, the solutions are accepted at a certain probability and will be calculated respectively in Adapted Simulated Annealing (ASA) process as the initial solutions. At last, conjugate gradient algorithm is utilized to search accurately and quickly. This genetic simulated annealing algorithm can effectively enhance the robustness of the algorithm and reduce the limitation of the searching area. The experimental results show that the proposed algorithm can calculate the thickness of 3-layer and 4-layer thin film within the range of 10 nm~5 μm and less than 1% calculation error

In view of the advantages of xenon lamp and its spectral energy distribution familiar to the sun, the xenon lamp is usually used to simulate the solar spectrum in the solar simulator. Because of the spectral differences between xenon lamp and the sun, it is necessary to develop a kind of minus filter to correct the spectrum of xenon lamp. The minus filter works under the irradiation of major light, so the materials must have low absorption and high anti-energy threshold. We prepare the film on quartz crystal by means of electronic beam and ion-beam assisted deposition system. Reasonable deposition mode and optimal processing parameters are adopted for a smooth surface layer with high surface quality and solidity, which passes through anti-energy test. The filter finally meets the applying demand perfectly.

The main error sources for measuring phase retardation of Wave Plate (WP) based on laser frequency splitting technology are analyzed. And the systematic errors, caused by the initial phase retardation of laser and misalignment of WP, are compensated. Error analysis shows that as to multi-order WP, the measuring uncertainty is 1.7′ and as to zero-order WP, 0.8′. Its measurement performances are tested in experiments, including comparing the measurement results when WP is aligned or inclined to the laser axis, measuring the phase retardation at different points among the WP aperture;and evaluating the misalignment angle between WP optical axis and its surface. According to experimental results, this system demonstrates the repeatability and reproducibility better than 0.02°, and the uncertainty better than 2′, viz. about λ/104.

For detection of the all-directional angle in the laser warning with the simple structure and stabilized performance, laser direction measurement method based on the polarization code optical fiber antenna was proposed. This method is to lay the polarizing code of the different polarizing eyepiece angle in the different direction. Then, each direction's optical fiber was coupled into a bunch of optical fibers. After the focusing collimation set of lenses injection solution leaning path of rays, the laser incidence direction was obtained through the judgment circuit analysis. In calculating and the simulating the polarizing laser energy attenuation, the solved electric circuit could examine the response electric current peak-to-peak value relation which was caused by the laser when the weakened polarization is biggest and the laser is weakest. The survey requirements were satisfied when the laser angular resolution is 30°. The experiment demonstrates that the smallest recognition voltage is 20 mV in each 20° division's polarization angle when original laser power is from 20 mV to 150 mV, which meets the requirements of all-directional laser direction survey.

A preliminary research on color simulation of stars was done to overcome the limitation of lack of spectrum information of stars in a star simulator nowadays. First of all, choose the suitable star catalog, decide the star number which would be researched on color simulation, and find the color index of the star. Simultaneously, deduce the functional relationship between the color index and the color temperature. And the color temperature of the star was decided. Then, calculate the radiation property of the blackbody under the color temperature. After that, according to the theory of colourometry, integrate the stimulus values of the blackbody and normalize the results to get the corresponding colourometry coordinates. Therefore, the color simulation of stars is implemented. Finally, explain the limitation of the research program. The experimental results show the feasibility of color simulation of stars.

Aiming at the drawbacks of the relative measurement, an absolute measurement technique using capacitive displacement sensor is presented in this paper. This technique is based on two coupling levels and multi-precision to achieve the absolute coding, which realizes the measurement of high accuracy and long-range. This paper gives a brief introduction of the structure of capacitive measurement system. And the principle of the system is also analyzed. Phase discriminator solves the problems that the accuracy of signals is decreased by the filter and long settling time is required.

According to the requirement of precision diameter measurement of the single crystal silicon sphere, the principle and characteristics of the measuring system were presented, and mathematical models were established for calculating the ellipticity of concentric interference rings with different beam incident angles. Furthermore, the diameter measuring errors of the fringe center were studied in different incident angle situations. It is shown that the error can be up to 6.6 nm when the original distance between inner face of the etalon plate and the adjacent silicon sphere surface equals 13.2 mm, and the incident angle equals 10-3 rad. A novel method was proposed for accurately adjusting incident angle of laser beam relative to the plate. The experimental results show that the incident angle can be less than 10-5 rad by applying this method.

According to the resource of error, the subpixel location error of the facula centroid has been classified as random error and system error. A simple and effective method is proposed to measure the location error of facula centroid. A measuring system which mainly consists of a high precision motorized translation stage of one dimension, POINT GREY Flea2-14S3 CCD camera and LED light source was designed and constructed. Based on a detailed analysis for experiment results, the periodic changing law of system error of LED facula centroid location in this measuring system was found. The results indicate that based on Flea2-14S3 CCD camera, the random error of facula centroid location is about 0.018 pixels, the system error is about 0.06 pixels, and the total error is about 0.063 pixels (1/15 pixels). According to the total error above, this CCD imager can be widely used in many measuring systems which count on facula centroid subpixel detection. The experiment shows that the measuring system is an effective means to estimate the location error of facula centroid.

The fiber-based Stimulated Brillouin Scattering (SBS) coupled equations were solved by the method of implicit finite difference with prediction-correction. The influences of the Brillouin gain coefficient, gain bandwidth and effective mode area on time delay and pulse broadening were investigated within the gain range of 0~16. The results indicate that the Stokes pulse will reach gain saturation at a larger gain for smaller gain coefficient, smaller gain bandwidth and larger effective mode area, and its maximum time delay is accordingly larger. At a given pump power whose gain parameter is in small signal range, the time delay and pulse broadening factor increase with the Brillouin gain coefficient increasing but decrease with the Brillouin gain bandwidth and the effective mode area increasing.

Layered Space-time code (LST) in free space optical communication system can greatly improve system transmission speed rate, but the corresponding error rate of the method impacts the reliability of system badly. Firstly, based on the analysis of optical intensity fluctuation, a Multiple Input Multiple Output (MIMO) channel model is proposed. Secondly, the decoding of LST is completed by using Serial Interference Cancellation (SIC) in free space optical communication, and then its performance is compared with that of linear decode algorithm. At last, the feasibility of the technique is proved by simulation analysis, which shows that in free space optical communication system, the error-rate performance by using SIC algorithm in LST is better than that of linear decoding algorithm.

Quantization is a very effective scheme for public digital image watermarking. Currently, there are lots of quantization-based public digital image watermarking methods. The quantization formulas used in these methods are in many different formats, whenas some formulas are better than the others and some formulas are equivalent. For the existing several quantization formulas in different formats, their equivalence is analyzed in strict theory and their optimality in statistics, and then the best quantization formula is obtained. Thus, current existing quantization-based image watermarking methods without the best quantization formula can be improved by using the best quantization formula.

To solve the problem of blurry edge in traditional linear interpolation algorithm, an improved algorithm is proposed. Firstly, we compute six interpolations of horizontal, vertical and diagonal direction with inverse distance square approaches in the neighborhood of the interpolation point, and then we define weighting factors with interpolation distance and direction gradient. Finally, we get the final interpolation by data fusion. Considering both the distance of interpolation and the direction gradient, the information of the edge and the texture of the original image are both effectively protected in this method. Results show that compared with the traditional bilinear interpolation approaches, the new one makes the mean squared error decrease, while the average gradient increases, which is effective to improve image resolution.

Since the compressed sensing video reconstruction frame by frame suffers interframe dithering, a multi-frame reconstruction algorithm of compressed sensing by group is presented. First, observe randomly frame by frame, and then make use of interframe correlation as three-dimensional signal. Second, in each step of the iterative process, update the reconstructed video by using the residuals. Finally, carry out the signal threshold processing based on three-dimensional transform, such as the 3D Dual-Tree Complex Wavelet Transform etc. Experimental results show that it can precisely reconstruct the original video by using randomly observed values, and achieve a higher signal to noise ratio. It is shown that algorithm uses the relevance of video frames effectively to eliminate the interframe dithering in reconstruction frame by frame.

With the development of online shopping, picture presentation as a shopping guide plays an increasingly important role in the online transaction. In many applications, it is very necessary to distinguish graphics and photographs before data mining. Based on the pictures of goods on the Internet, several ways are discussed to distinguish graphics from photographs, and their limits are analyzed. Then, a simple and effective algorithm is proposed. Firstly, text-images were divided from photographs by human visual habits, and token as graphics. Then, according to the analysis of dominant color and color diversity of the left images, the part generated by computer was figured out. The result on an online goods picture library including 40 000 pictures demonstrates the effectiveness and applicability of the algorithm.

For NeighShrink method used in the image de-noising, a new image de-noising algorithm is proposed to keep image edges more effectively, and it mainly improve the domain of NeighShrink which is fixed. The new method can segment the image into many domains adaptively to de-noise the images. Furthermore, combined with wavelet correlation in the same layer, we get various irregular neighborhoods with a fixed window, and choose the coefficients which have closer geometric distance and are in the same irregular neighborhood to improve NeighShrink. This method decomposes noisy images with stationary wavelet transform to keep phase invariance. Then, in accordance with special rules, it segments the low frequency sub-band by using Pulse Coupled Neural Networks (PCNN) model, and then gets the approximate information. And the edge information will be protected during the de-noising process. A better restoration of images is demonstrated in the results of experiments, with detail of images kept as well as image noises decreasing.

Aiming at the often occurred “buried alive” accidents in the coal industry, bulk cargo docks etc., a alarm system was designed by using structure of IRCCD+CPLD+DSP. Real-time capturing, processing and displaying of infrared image were implemented to recognize the life target in tipper consequently. In the image capture module in which CPLD is the controlling core, SAA7111A video processing chip was used to implement changing analogue image to digital image and synchronously controlling time-sequence. In the image processing module in which fix DSP TMS320C6416 is the core, infrared image was enhanced and the image feature was extracted, and life target recognizing in tipper was realized. Sound-light alarm delay circuit in which 555 is the core is used to implement alarming. In the ending of the paper, the result of feature extraction at the different interior thresholds was given. Experiment results demonstrate that the ideal result can be achieved at the 0.8 interior thresholds of gray level.

For binary optical elements, diffraction efficiency is one of the important performance indexes. Under the oblique incidence cases, the optical path difference of light beam distributed on the edge of steps can be calculated by using the geometric method. Then, according to the multi-slit Fraunhofer diffraction, the expression of diffraction intensity of reflection step grating is deduced. Under the two limit conditions that the number N of different steps changes from 2 to infinity, we discuss the expression of light intensity’s change. In general case, we deduce the expression of diffraction efficiency which is different from that of other articles. Experimental results show that the step number and ratio between the etching depth and the light wavelength are closely associated with the diffraction efficiency of reflection step grating.

An active optical fiber loop is consisted of a low noise EDFA, an optical band pass filter with 0.8 nm pass band, and 10 439 m standard SMF fiber. The detailed test was performed and good test results were obtained, which indicated that the active optical fiber loop has a good ability of signal pulses replicating. Based on the active optical fiber loop, an optical buffer unit was designed. The principle of the optical buffer was explained and the definition of unit storage time and maximum storage time were given. Experiment results indicate that the optical buffer unit can restore the input optical pulse signal. It has a minimum storage time nearly 50 μs, and the maximum optical pulse signal replicators are above 500. In another words, the maximum storage time is nearly 25 ms.

In order to adapt the optic intensity variation of the environment and to obtain stable and high quality image features, an automatic control algorithm on CCD camera electric shutter is proposed. This algorithm mainly consists of two parts: auto shutter control (ASC) based on cubic interpolating polynomial of common logarithm of the image average gray and image grey stretch technique based on experiment data analysis. The experimental result shows that the proposed algorithm has fast and stable convergences, and those captured images by using this technique have good quality and adaptability at different environment illumination.

To form a new kind of Plasma Photonic Crystals (PPCs) filters, a plasma defect layer was filled in the PPCs. The effective refractive index can be changed by alternating the electrical parameters of the defect instead of changing the dimension of PPCs, which makes the resonant frequency in plasma defect layer offset. Then the filter frequency can be tuned. A Finite-difference Time-domain (FDTD) algorithm is applied to calculate the transmission spectrum of the plasma photonic crystals. The result show that the tunability of the PPCs filters can be achieved by changing the parameters of the plasma defect layer which changes the resonance frequency.