Computer Vision Method in Human Motion Detection

Human motion detection based on computer vision is a frontier research topic and is causing an increasing attention in the field of computer vision research. The wavelet transform is used to sharpen the ambiguous edges in human motion image. The shadow＇s effect to the image processing is also removed. The edge extraction can be successfully realized. This is an effective method for the research of human motion analysis system.     

Real-time Extraction of Target Position Information Based on Bi-orthogonal Wavelets

Extraction of flying target position information is the prerequisite for passive infrared guided missiles to track the target. The existing missile detection system senses the target＇ s infrared radiation, and then the generated signal is sent to signal processing circuits for extracting the target position information. In order to improve anti-interference capacity of the detection system, an algorithm of module maximum edge detection based on the bi-orthogonal wavelets is proposed to replace its hardware. The signal can be decomposed in one layer, only its high frequency detail is reconstructed. After some calculations, the average target position can be obtained. The algorithm＇s real-time implementation with DSP is also discussed. To reduce the execution time, the program structure can be optimized with double buffers in memory. This im- plementation is verified by simulations. The results show that the method has only a small amount of calculations, can meet the requirements for finding out the target position in real-time and needs not the traditional processing circuit.     

Processing Defect Analysis and Nondestructive Evaluation Technology for Si_3N_4 Bearing Ball

Researches on the processing method of ceramic bearing ball,the formation and propagation of defects in the manufacturing and the nondestructive evaluation（NDE） are summarized in this paper.The key for successful processing of high strength ceramic balls is to avoid producing related defects.Many investigations show that the material microstructures,defects as well as mechanical processing parameters influence the final surface quality significantly.Most of NDE technologies,such as radiation,ultrasonic,dye-penetration and laser scatter,have been studied for ceramic bearing ball surface inspection around the world.So far,the difficulties to develop the perfect NDE system for ceramic bearing balls,which are caused by the defect variety and surface unfolding,have not been overcome yet.     

Bark-Wavelet Analysis and Hilbert-Huang Transform for Underwater Target Recognition

Recognizing the underwater targets by the radiated noise information is one of the most significant subjects in the area of underwater acoustics. Based on the theory of auditory perception, a novel recognition approach which consists of the algorithms of Bark-wavelet analysis, Hilbert-Huang transform and support vector machine is proposed. The performance of the proposed method is validated by comparing with traditional method and evaluated by the recognition experiments for SNRs of 0 dB, 5 dB, 10 dB, 15 dB and 20 dB. The resuits show that the average recognition rate of the method is above 88% and can be increased by 0.75 % to 6. 25% under various SNR conditions compared to the baseline system.     

A Sub-pixel Image Processing Algorithm of a Detector Based on Staring Focal Plane Array

Optical micro-scanning technology can be used to increase spatial resolution of many optical imaging systems, especially thermal imaging system. One of its key issues is relevant image processing algorithm. A fast reconstruction algorithm is proposed for two dimensional 2 × 2 micro-scanning based on the sub-pixel imaging and reconstruction principle of two-dimensional staring focal plane arrays （FPA）. Specifically, three initialization methods are presented and implemented with the simulated data, their performances are compared according to image quality index . Experiment results show that, by the first initialization approach, timely over-sampled image can be accurately recovered, although special field diaphragm is needed. In the second initialization, the extrapolation approximation in obtaining reconstruction results is better than either bilinear interpolation or over-sampling reconstruction, without requiring any special process on system. The proposed algorithm has simple structure, low computational cost and can be realized in real-time. A high-resolution image can be obtained by low-resolution detectors. So, the algorithm has potential applications in visible light and infrared imaging area.     

Conversion of Torpedo＇s Lake and Sea Test Data Based on Grey System Theory

As only a little information can be obtained from torpedo＇s lake and sea tests, and the torpedo＇s life does not distribute typically. If conventional methods are used to convert the environment factor for torpedo＇s lake and sea tests, their results can not reflect the actual conditions. A conversion model of the environment factor for torpedo＇s lake and sea tests is set up based on the GM（1,2） model of the grey system theory. For the merit of the grey system, the problem of uncertain life distributions and few samples can be solved. The calculation results show that the method is easy, realistic and high precise.     

Structural Characteristics and Decomposition Mechanism of 2, 2＂-Dimethyl-5, 5＇-Azotetrazole

The geometry and the potential curve of thermal decomposition for 2, 2＇ dimethyl-5, 5＇-azotetraol （2-DMAT） are calculated by ab initio quantum chemistry method. The structural characteristics and decomposition mechanism are carefully studied. It is found that the terrazzo ring satisfies 4n ＋ 2 rule and it is a conjugated π-systems for 2 DMAT, The azotetrazol has aromaic characteristic and its thermal decomposition can proceed in two steps： ring opening and N2 separation, The activation energies of the two steps are 152.3 kJ/mol and 44.67 kJ/mol respectively. The ring opening is the rate-controlling step.     

Study on System Failure Probability Model Based on Dynamic Fault Tree

Aiming at the characteristics of complex logic relation and multiple dynamic gates in system, its failure probabil- ity model is established based on dynamic fault tree. For the multi-state dynamic fault tree, it can be transferred into Markov chain with continuous parameters. The state transfer diagram can be decomposed into several state transfer chains, and the failure probability models can be derived according to the lengths of the chains. Then, the failure probability of the dynamic fault tree analysis （DFTA） can be obtained by adding each chain＇ s probability. The failure probability calculation of DFTA based on the continuous parameter Markov chain is proposed and proved. Given an example, the analytic method is compared with the conventional methods which have to solve the differential equation. It is known from the results that the analytic method can be applied to engineering easily.     

火炮炮膛质量检测技术研究

介绍了一种利用ＣＣＤ摄像镜头来获取火炮炮膛表面疵病信息的检测装置，论述了该装置的设计思想和实施方案，该装置以ＣＣＤ摄像镜头为传感器，使用ＩＢＭ－ＰＣ／ＡＴ４８６机作主控和图像处理系统，主要用于各种火炮炮膛表面疵病（烧痕，裂纹，瑕点等）的实时自动检测，也可对石油，化学工业中各种金属管道进行自动检测。     

微小尺寸零件表面缺陷光学检测方法

为实现微小尺寸零件表面缺陷的快速精确检验,对微小尺寸零件表面缺陷光学检测技术进行了研究.以典型微小尺寸环形零件为研究对象,从零件的表面特征、缺陷类型及检测算法出发,建立了基于机器视觉的微小尺寸零件表面缺陷光学检测系统;通过对微小尺寸环形零件表面缺陷纹理特征研究,提出图像梯度方差加权信息熵算法抑制表面随机纹理,提取微观表...     

基于机器视觉的枪弹钢芯自动检测系统

针对现有的手工枪弹检测技术效率低精度差且自动化程度不高的问题,提出一种采用机器视觉技术设计的枪弹钢芯检测系统。该方法从钢芯传送定位、对钢芯最佳照射、选择图像采集设备和图像处理和识别着手,对钢芯尺寸、底面缺陷、表面缺陷进行检测,经过平滑去噪和图像增强等预处理后,进行尺寸测量和缺陷识别,并通过人工复检证明了准确性。实验结果证明：相比传统的人工肉眼检测,该自动检测系统所耗费时问大幅度缩减,检测效率和检测精度均有较大提高。     

基于数字图像处理的经纬仪对心测量方法

以经纬仪对心为研究对象,使用CCD传感器采集位置图像并经图像处理测量出经纬仪对心偏差,驱动二维高精度电控位移台移动来实现自动对心。对经纬仪原光学对心镜系统进行了改造,研究了图像处理的方法,包括图像预处理和圆标记中心定位算法,采用了混合滤波和Sobel-Zernike亚像素边缘检测方法。仿真实验表明：基于图像处理的经纬仪对心测量方法弥补了对心需要反复操作的缺点,大大缩短了测量时间,测量精度也可以达到5.5μm。     

轴承钢球表面缺陷的快速检测方法

针对目前轴承钢球表面缺陷检测方法中存在的问题，提出了一种快速检测方法。设计的钢球表面展开结构，解决了钢球表面快速展开问题，能够在快速展开的同时保证检测面的完整性；通过改进光源和检测槽设计，消除了图像采集过程中的钢球表面镜面反射干扰；结合计算机数字图像处理与模式识别技术，实现了背景和钢球的分离、钢球和疵斑的分离，达到了获取钢球表面缺陷特征的目的，使得检测过程自动化。实际使用表明，该方法能够对钢球表面的点子、擦伤、烧伤、划伤、裂纹等缺陷进行检测，可用于钢球表面缺陷在线检测。     

微小零件几何尺寸自动检测装置

提出一种基于图像处理技术完成微小零件几何尺寸的自动检测装置的实现方案。详细介绍了该装置的控制过程、检测原理和实施途径。通过采用CCD数字图像处理技术，实时采集被测零件的外形和表面图像，通过特定的算法在计算机上完成对微小零件几何尺寸的测量；通过采用可编程逻辑控制器（ProgramLogicControl，PLC）技术，实现对微小零件的输送、定位、检测、抓取等环节按一定时序有节拍的进行控制。经对工件实际测试和结果分析表明：该装置可以实现实时的连续自动检测，显著提高检测效率，检测的分辨率可达到4μm以上。     

基于下视图像的打击效果评估系统研究

研究了打击效果评估的基本概念及该技术在国外的发展与应用情况,提出了基于下视图像的毁伤特征提取方法与典型目标毁伤评估准则。从物理级、功能级和系统级三个层次对打击效果评估进行了描述,提出了基于变化检测的物理级毁伤效果评估方法、基于自动目标识别的物理级毁伤评估方法和基于专家知识库的功能级毁伤评估方法,开发实现了基于下视图像的典型目标打击效果评估系统。仿真结果表明,所提出打击效果评估方法是有效的。     

基于计算机视觉的人体运动目标检测

基于计算机视觉的人体运动目标检测是计算机视觉领域中倍受关注的前沿课题。本文采用小波变换的方法,有效地处理了普通CCD摄像机所采集的人体运动图像的模糊,同时,去除了阴影对图像处理的负面影响,较好地提取出基于标志点的人体运动图像的边缘,为人体步态分析系统的研发提出了一个切实可行的方法。     

一种基于图像检索的枪弹识到系统

针对国内枪弹识别技术的现状和刑事工作人员的需要，设计了一套枪弹自动识别系统。该系统能够完成枪弹痕迹图像的自动采集，对采集到的图像进行预处理使其规范化，规范化图像存档于数据库中。利用基于多小波直方图的图像检索方法，在数据库中检索与目标图像最相似的样本图像，供枪弹检验人员作进一步的决定。系统所具有的自动采集、规范、存储、检索等功能，解放了显微镜下逐一比对、查找这种繁琐而耗时的人工方式，鉴定速度和准确度大幅度提高。