【作者】 韩磊；

【导师】 张海洋；

【作者基本信息】 北京理工大学 ， 电子科学与技术， 2015， 硕士

【摘要】 在安保、反恐以及现代化战争中，遏制训练有素、装备精良且埋伏在隐蔽地点的狙击手是一项非常艰巨的任务。这些狙击手通常具有很强的攻击性，他们的存在严重威胁着出现在公共场所的重要人物和在战场作战士兵的生命安全，因此反狙击装备的研究受到越来越多的重视。目前，反狙击装备的研究主要基于三类技术，分别是声探测技术、红外探测技术和光电探测技术。采用前两种技术时，只能在狙击手采取行动以后才能将其发现。为了实现对狙击手先发制人，基于光电探测技术，依据“猫眼效应”原理提出了一种基于脉冲激光主动探测狙击手位置的新方法，并围绕探测系统的探测机理分析，目标图像识别算法和嵌入式信号处理等主要部分展开研究。首先，根据光电系统猫眼效应的原理，即入射到瞄准具分划板的激光束会沿原路返回，且反射强度比漫反射波高102～104倍，对猫眼探测系统的一般结构进行了详细介绍。创新性的提出在探测系统中引入脉冲激光进行探测，使用主被动图像相结合的方法来降低虚警率，并通过对比几种主被动图像获取方式，确定了通过GPIO控制激光器配合图像采集来获得主被动图像的方法。其次，研究了猫眼目标识别的图像处理算法，主要包括抑制图像噪声，二值化处理，形态学处理以及基于可疑连通域面积的不同采用不同的目标识别准则等内容。然后，对信号处理系统的软、硬件结构进行了详细的分析、讨论和设计，主要包括信号处理板各模块原理和选型原则，PCB设计及相关技术，嵌入式操作系统DSP/BIOS的运用和底层驱动软件的设计。最后，对系统进行了外场实验，将目标分别置于300m近距离和800m远距离处，并对两个典型距离的实验结果进行了详细分析。实验结果显示基于脉冲激光的狙击手主动探测系统在保证准确探测的同时具有低虚警，证明了系统的有效性和先进性。更多还原

【Abstract】 In terms of security, counter-terrorism and modern warfare, preventing snipers, whoare well-trained, well-equipped and always ambushing in a hidden place, is a very difficulttask. These snipers are usually aggressive and always threat the lives of the VIP and thesoldiers that fighting in the battlefield. Therefore, the country gives more and moreattention to the research of anti-sniper equipment.Currently, the research is mainly based on three technologies, acoustic detectiontechnology, infrared detection technology and photoelectricity detection technology. Whenusing the first two techniques, only after the snipers acting could we find them. In order toachieve pre-emptive strike against snipers, the paper based on the photoelectric detectiontechnology and the principle of cat eye effect, propose a new method to detect the positionsof snipers. The study is conducted and around three main parts, for example, the opticalsystem of cat eye detection, target recognition algorithm and embedded signal processing.First, according to the physics principles of cat eye effect, the laser beam that incidenton the gunsight reticle can return along the same route and the intensity of reflection is102to104times higher than that of diffuse reflection, the general structure of cat eye detectionsystem was described in detail. The paper proposed the introduction of laser pulseinnovative, and used a combination of active and passive image to reduce the false alarmrate. By comparing the access of several active and passive images, the study determined toget active and passive image by laser that controlled by GPIO and image acquisition.Secondly, the paper studied image processing algorithms of cat eye target recognition,including image noise reduction, binary processing, morphological processing and differentcriteria for different target identification which based on the suspicious connected domainarea.Thirdly, the paper analyzed and discussed the architecture of software and hardwareof the signal processing system, including the principle and selection of the signalprocessing board modules, PCB design and related technologies, the use of embeddedoperating system BIOS and the design of the underlying driver software.Finally, the system has been tested by outside experiments that the objectives wereplaced in300m at close range and800m at far range, and the experimental results of twotypical distance were analyzed in detail. The results showed that the new developed sniper detection system could ensure detection accuracy and had no false alarm at the same time,therefore, the effectiveness and advancement of the system was demonstrated.更多还原