The dim photoelectrical target in space has the weak signal energy. If the target motion parameters and brightness information were lack in the low Signal to Noise Ratio (SNR) conditions, it would be difficult to achieve the detection and tracking of dim target. Due to the character of particle filter in the signal processing, its application in the dim target tracking was researched. First, the basic theory of particle filter in the application of the tracking was analyzed. For the tracking of target with unknown luminance, a method for constructing the objective function via point spread function was presented, and the weight of the particles could be updated. Then, by using the circular orbit as the circulating model of target, the speed range of the space target on the sensor was estimated, thus optimizing the initial conditions for the particles. Finally, a simulation experiment for tracking the dim target was made to testify the effect of this method on decreasing the number of particle.

Aiming at the feature of Electronic Support Measures (ESM) detecting the bearings and Doppler frequencies simultaneously, a target tracking algorithms is proposed which uses ESM’s Doppler frequency and bearing measurements based on Modified Gain Extended Kalman Filter (DB-MGEKF). Compared with bearings-only target tracking algorithm, DB-MGEKF increases the Doppler frequencies measurement and introduces the MGEKF which can handle the nonlinear system preferably. The method improves the accuracy of target estimates as well as the stability of the filtering. Moreover, the observer’s maneuvering can be avoided. Monte-Carlo simulations with result analysis further illustrate the effectiveness of the algorithm.

In consideration of the requirement and structure of the practical turntable, the indispensable research on the force of friction is analyzed, which impacts the speed stability performance. A kind of control strategy based on the Variable Structure Control (VSC) is introduced. According to the invariant features and the exponentially approaching rule of the VSC, the new control strategy is designed by the application of traditional double closed loop control system and VSC system. According to the simulating results and the debugging results of the control strategy on the practical turntable system, we can validate that it agrees to the simulating results, which demonstrate the effectiveness of the control strategy based on VSC in improving the speed stability performance.

Rotating target is the common equipment used to test tracking performance of photoelectric theodolite indoor. With the aim to study the impact of rotating target characteristics on target tracking performance of photoelectric theodolite, we introduced the basic structure of rotating target and the core parameters which can impact the motion features of rotating target. Using the method of spherical trigonometry, we established the coordinate equation for the angle of azimuth and elevation. According to the coordinate equation, we presented how the parameters of rotating target take responsible for the target motion features. Finally, we demonstrate the difference of motion features between rotating target and actual tracking target; suggesting that using rotating target to test the tracking performance of photoelectric theodolite is inaccurate.

In order to improve the sensitivity of the CCD, the makers add microlens array on the CCD surface to increase the quantum efficiency. Because the microlens can focus the light ray, the rays’ incidence angel must be small. Normal lens decrease the incidence angle in image space by increase back focal distance, but the overall length increases and the imaging effect is not better. An aerial digital camera lens with telecentric beam path in image space is designed; the chief ray exit angle of the edge fields is 0°.In this way, the special requirements for lens of digital camera with microlens array is solved. It is proved that the telecentric beam path camera lens has high resolution, good illumination homogenization in image plane, short overall length and light weight.

A modified voting star recognition algorithm with main star triangle angular determination scheme was proposed. For better robust performance, mode vector can be generated with three stars in the new star recognition algorithm. After the selection of stars according to certain rule, the angle at the star to be recognized in the triangle that is composed of every selected star triplet was calculated, and then act as mode vector for voting recognition. Because of higher dimension in generation and wider volume distribution, the algorithm could provide better anti-noising capacity and performance in recognition success rate. Simulation and result analysis proves the effectiveness and certain value of algorithm.

For compensating the defocus caused by the changes of temperature, pressure and other environmental parameters of an aerial camera in the air, an autofocusing system for aerial camera based on opto-eletronic auto-collimation method was designed. Utilizing raster with high contrast in the camera, this system avoided the situation that autofocusing based on image processing technology failed when object is with low contrast. Modulated by raster in object space, image space and the swing of reflector, the signal of light intensity was easy to be processed. Combined coarseness with fine searching strategy, the system guaranteed the validity of adjustment result. Experimental results show that the autofocusing process is accomplished in less than 20 s, and the repeatability is less than half focus depth, which satisfies the focusing requirement of aerial camera.

A novel method for determining weight of hybrid chirp in ultrashort laser pulses is presented based on asymmetry of Interferometric Autocorrelation (IAC) envelope, which is highly sensitive to weight of hybrid chirp and chirp magnitude. By analyzing and calculating the asymmetry of IAC envelope, the asymmetry envelope to determine weight of hybrid chirp, and center halfwidth of IAC envelope ratio and peak value of IAC envelope difference as a function of hybrid chirp for various weights were obtained. The theoretical analysis is supported by experimental ratio of IAC envelope and difference of IAC envelope. With this method, the hybrid chirp in ultrashort laser pulses was found and the linear and square chirp in chirped ultrashort laser pulse was measured 0.20 and 0.80, respectively. The measurement about weight of hybrid chirp in ultrashort laser pulses is realized.

In order to detect obscure targets by the use of LAser Detection and Ranging (LADAR), we constructed the LADAR echo model of obscure targets and obtained the simulation data of LADAR echo of obscure targets. Levenberg-Marguardt method was used to fit the echo waveforms, which decomposed the waveforms into series of Gaussian components, and obtained the number and the position of the Gaussian components, half-width and amplitude of each Gaussian component. Echo components of sub-target in different channels were obtained by time matching of separated sub-echoes respectively, and the characteristics extraction of various targets has been implemented. Classifying the echo characteristics by using K-means helps to classify various echo components. An anticipated result of simulation is obtained by using this method.

According to the optical activity of oil mist, the method of oil mist concentration detection based on optical rotation effect was presented. Basic principle of oil mist concentration detection system was introduced. The structure of detection system was designed. Mathematical model of detection system was established. Simulation and experiment research on the detection system was carried out. In oil mist chamber, linearly polarized light which is used as input optical signal is produced by polarizer and detected by analyzer. Using single-light-path detection method, detection system is compensated and modulated by Faraday modulator and calibrated by weighing method. Concentration information is obtained by photoelectric conversion. The experimental results show that the method has good linearity and stability, and detection effects are good when oil mist concentration is low.

The calculation method of conventional gasdynamic CO2 laser is always based on relaxation dynamics. In the paper, energy transfer processes in donor nozzle of mixing gasdynamic CO2 laser(MGDL) are treated as chemical reactions and freezing efficiency of N2(v=1) vibrational level is studied by using method of chemical reaction kinetics. Using forward reaction rate constants given by other papers, the backward reaction rate constants are calculated. Distribution of N2 (v=1) molecules in donor nozzle of MGDL and influences of total pressure and molar fraction of H2O on freezing efficiency of N2(v=1) vibrational level are studied. The results indicate that decrease of mass fraction of N2(v=1) molecules happens mainly in and around the nozzle throat. Compared with the case of not containing H2O, the freezing efficiency of N2 (v=1) vibrational level decreases remarkably while containing H2O. When total pressure increases, the freezing efficiency of N2(v=1) vibrational level decreases, but just a little. Comparatively, with increasing of molar fraction of H2O, the freezing efficiency of N2 (v=1) vibrational level decreases rapidly.

A camera self-calibration technique based on factorization and bundle adjustment is proposed. With the hand-held and fix-focus camera undergoing at least three arbitrary motions around the calibration pattern, all the intrinsic parameters and the distortion coefficients can be obtained. The proposed method has three novelties. Firstly, its robustness is markedly increased since all the images are aligned in the factorization process. Secondly, the non-linear optimization algorithm bundle adjustment guarantees high accuracy. Thirdly, the proposed method does not require specialized calibration pattern or rigid camera motion, which makes it be used in a wide range of applications. Both simulation and real images experiments proved the feasibility and applicability of the proposed method, particularly applying to those close-range photogrammetry system based on image sequences. As a result, the new algorithm has been successfully applied to the feature point measurement system TN 3DOMS.FP v1.2 with independent intellectual property right of Saint Buffalo Technology Co., Ltd.

Electric field enhancement in single layer multimode guided-mode resonance filters (GMRFs) is presented. The physical basis that a single layer GMRF can be functioned simultaneously as the waveguide and the phase matching element is demonstrated. Multimode resonance occurs as the grating thickness increases, resulting in electric field enhancement of high order leaky modes. The order of the electric field of maximum amplitude indicates the leakage level of the GMRFs. Thus the bigger maximum amplitude of the electric field will generally correspond to the smaller filter linewidth at the same resonance wavelength.

A leak detection and location system for SF6 based on infrared image is designed and developed. By using the high speed microprocessor TMS320C6416 and its configurable video ports and combined with the powerful control capability of the S3C6410, the real-time infrared image where SF6 leakage may occur is collected and processed. Combine with the frame of difference algorithm and the local entropy algorithm, detect and locate leaks. Experimental results show that the algorithm not only effectively improves the precise detecting location of frame difference algorithm and reduces calculated amount of local entropy method, but also effectively reduces the effect of diagnosis process with object shape, surface defects and environmental background noise etc, which can efficiently and accurately implement SF6 leak detection and location.

The optical system with N-type fiber (patent pended) was constructed, and a new fiber-type fluorescence analyzer of PKU was developed for overcoming the disadvantage of traditional PKU, which could not monitor the exciting light intensity fluctuations and made the measured fluorescence data have greater error of the instability. According to the U.S. Committee for Clinical Laboratory Standards (NCCLS) documents, American Society for Quality (ASQ) of the QS-9000 standards and related literatures, a quantitative evaluation of the related equipment indexes was presented by computing between fluorescence analyzer and the national standard equipment (THERMO Varioskan Flash). Using the standard, the results show that the indicators of fluorescence analyzer are in line with requirements, and then point out the direction of the instrument to be improved.

In industrial Ethernet, large volume of data is transferred by UDP, but the traditional UDP is non-linked and don’t deal with the dropout packet and delay. A new protocol Tolerate Error Reliable UDP (TERUDP) is put forward based on traditional UDP. It re-defines packet loss. Moreover, the concepts of valid delay and invalid delay are introduced, according to which it determines whether to retransmit missing packet and the delivery time of retransmission application. The paper suggests detecting network congestion according to virtual receiving rate and actual receiving rate and packet loss rate and then responding to it. Simulation results show that TERUDP without changing the efficiency of traditional UDP improves the reliability, reduces the network packet loss rate and delay, and increases the network data transmission efficiency.

The method, by using photoconductive semiconductor switch to generate the offset voltage for terahertz time-domain spectrum system, was adopted to improve system’s performance. Based on the analysis of the terahertz wave’s generation mechanism by photoconductive semiconductor switch, the problems within the current terahertz time-domain spectrum system were fully examined, the solutions to solve the contradiction between the mechanic noise induced by chopper and the photoconductive antenna’s offset voltage, were discussed. Last, after testing the characteristic parameters of photoconductive semiconductor switch, a scenario which took advantage of high-voltage direct current source, photoconductive semiconductor switch, and nanosecond laser to generate 0 to 9 kV offset voltage without mechanic noise, was put forward to level up the amplitude of the offset voltage to improve terahertz time-domain spectrum system.

To keep micro projection system compact and highly efficient, an illuminating system is proposed based on LCoS and LED. This system follows the principle of Kohler illumination. By modifying Kohler illumination system and setting the rational stop, uniformity lighting zone can be got on the LCoS plane. Besides, this system becomes simple and compact. Spherical lens and aspherical lens are used in this illuminating system respectively. And simulation results show that illumination efficiency is 86% for spherical lens and 95% for aspherical lens. When using aspherical lens, transmission efficiency of the whole projection system is 9.94% and spatial uniformity is above 90%. The thickness of optical module is less than 9 mm.

Optical communication system requires the high isolation and low polarization sensitivity of Tunable Optical Filter (TOF). At present, the most of TOF based on thin film are hard to meet this requirement. The reason of TOF transmission spectrum degradation is analyzed by simulation. Referring to the center wavelength and bandwidth separation when different polarization beam in tilted incidence, an optimization design is proposed based on symmetric optical path. It improves the isolation of TOF to more than 30 dB, and restrains the polarization dependent loss within 0.2 dB. Meanwhile, different polarization beam have the same optical path because of the symmetric optical path, which can avoid adding additional wave plate. The module performance is tested, which realizes that 96 ITU-T channels can be tuned in the range of C band. The experiment results are in good agreement with the analytical ones.

Scale factor nonlinearity in an open loop Fiber Optic Gyroscope (FOG) indicates the maximum inaccuracy of rotation rates within the whole measurement range. Because of the non-linearity of the phase modulation, scale factor non-linearity of open loop FOG is inherent and will become worse if the bandwidth of preamplifier is limited. This paper educed theoretically the facts that will make the nonlinearity of scale factor even worse when all harmonic components are amplified dissimilarly. Finally, experiment indicates that 20 multiples of the base frequency bandwidth of a preamplifier with a 1st-order RC low-pass filter is needed if scale factor nonlinearity we wanted is less than 500 ppm or better.

According to the nonlinear polarization rotated effect, a self-starting mode-locked ring cavity fiber laser is obtained and optimized through changing the fiber length and coupler parameters. As the seed of an all-fiber ultra-short pulse amplifier with master oscillator power amplifier (MOPA) structure, the pulse which has 2.042 W in average power, 1 063.8 nm in wavelength, 41.3 MHz in repetition frequency, 45 ps in impulse duration, 49.4 nJ in single pulse power and 1.1 kW in peak power, is realized, taking double-cladding fiber as the gain medium. The optical conversion efficiency of the whole system is 40.8%, which could be supportable to the higher power all-fiber pulse laser.

To remove impulse noise from image before some other process, an adaptive switching filter based-on neural network noise detector is proposed. This method describes each pixel with pixel value and its neighborhood characteristics and takes these as inputs of the neural network to identify pixels which are likely to be contaminated by noise with the trained neural network automatically. According to the idea of switching filter, the noisy pixels detected are processed by mean filter with adaptive window size, and only noise-free pixels of the window are involved in the average computation. Compared with some other common filters, the experimental result shows that this BP neural network has high accuracy of salt and pepper noise detection. Besides, this filtering process is superior in denoising effect, details preserving and time consuming reduction without manual intervention.

A compression algorithm based on 3D-DWT and listless 3D SPIHT in conjunction with Band of Interest (BOI) for hyperspectral images is presented. It is specially developed for the LASIS (Large Aperture Static Imaging Spectrometer) system according to its interference properties. Firstly, decompose the hyperspectral image sequences with asymmetric uniform-tree 3D-DWT. Secondly, protect its major hyperspectral coefficients by BOI method. Finally, give a modified listless 3DSPIHT algorithm to encode the transformed images. The numerical experiment results show that the Peak Signal-to-noise Ratio (PNSR) is more than 40 dB at 8:1 compression rate; and achieves an efficient protection of hyperspectral information.

For the fusion of multi-focus images of the same scene, a novel algorithm is proposed based on the multiscale products in lifting stationary wavelet domain. The selection principles of the low frequency subband coefficients and bandpass subband coefficients are discussed respectively. When choosing the low frequency subband coefficients, we present a scheme based on a new sum modified-laplacian combined with the selection and weighted scheme. When choosing the high frequency subband coefficients, local modified Laplacian is proposed based on multiscale products according to enhancing edges structure of multiscale products while weakening noise. Then, we present a selection principle based on the local modified Laplacian. The experiments show that the algorithms proposed in the paper can not only extract all the useful information of the source images and transfer to the fused images, but also effectively restrain the noise influence. Compared with the traditional methods, a better performance is obtained in terms of both visual qualities.

The vibration and friction factors of two-axis motion platform cause video jitter and because of image feature extraction and matching influenced by illumination variation, noise and so on, the wrong motion vector will be estimated. The multi-scale space two dimension high precision video stabilization algorithm is proposed based on concentration degree. Invariant features are extracted, and then the depth first nearest neighbor search algorithm is used to get the matching double points. At last, concentration degree is computed, and different computing models are built to estimate compensation parameters by optimal concentration degree principle. 320×240 pixels gray video sequence is used in experiment, and the algorithm proposed in this paper has high precision. The estimation time is just 13% of Random Sample Consensus (RANSAC), which improves the performance of video stabilization.

A complete orthogonal discriminant locality preserving method based on minimal criterion is proposed on the theory of unsupervised discriminant projection. According to the space information among samples belonging to the same class, the proposed method redefines the within-class and between scatter matrix. Then, according to the objective function of unsupervised discriminant projection, the new objective function is derived, which can address the small sample size problem by being projected into total scatter matrix non-null space. The algorithmic procedure of the proposed method based on QR decomposition is given. Finally, experimental results on face database demonstrate the effectiveness of the proposed method.