主被动光学联合检测算法研究

【作者】 张鹏；

【导师】 李晓峰；

【作者基本信息】 电子科技大学 ， 光学， 2007， 硕士

【摘要】 随着各种新材料和新技术在军事上的应用,军事目标的可检测性不断减弱,目标的机动性不断增强。以往由单一探测手段、单一探测器件的检测数据已远远不能满足需要。随着探测技术、探测手段正不断的进步和数据融合技术的不断成熟,目前联合探测逐渐成为主要的探测手段,并广泛的应用于军事和民用领域。联合探测手段可以实现对目标更精确的探测,在很大程度上弥补了单一探测手段的不足。本文将联合探测技术引入了光学探测系统中,采用主动和被动两种方式,利用不同的光波段来对目标进行探测。这样可以提高系统的探测精度。本文进行了光学联合探测系统的目标特性和背景特性的研究。重点分析了主动探测波段的飞机目标的激光雷达散射截面(ORCS:Optical Radar Cross Section)和3～5μm和8～12μm波段的红外辐射特性,以及对系统选取波段造成影响的背景的特性。通过分析可以看出,针对喷气式飞机目标,采用8～12μm波段进行被动探测,无论是从其自身辐射特性还是其传输特性都要优于3～5μm波段。对光学联合探测系统的构建提出设想。对系统的子系统进行了分析和研究,主要分析激光雷达子系统和长波红外子系统的关键技术和主要的影响因素。对光电信号处理方法进行研究。对如何提高光电信号的检测概率进行了重点分析。通过分析可以看出,为了提高各子系统的检测概率,应进行积累检测。特别是对微弱的点目标,在红外波段采用“软判决”的检测方法,在充分利用目标的运动信息的情况下,可提高目标的检测概率。重点进行了光学联合检测算法的研究。针对联合探测系统的具体情况,主要应用了Neyman-Person检测融合算法和秩K检测融合算法对系统进行分析,通过分析看出,本系统在两被动探测子系统的检测概率相差较大时,两种融合算法都有自身的缺陷。本文对秩K检测融合算法进行了相应的改进,并对改进后的算法进行了分析和仿真。通过分析可以发现,改进后在虚警概率增加不大的情况下,总的检测概率却有很大的提高,同时可以减小系统融合中心的运算量和融合所需的时间,增加了融合的实时性。更多还原

【Abstract】 With the new material and new technology applying in military, the detectability of military target is lowing, and the mobility is increasing. The information coming from one detector by a single detecting means has can not satisfy the need. With the improving of new detecting technology and new detecting means, also with the improving data fusion, multi-detection has being the main one detecting means. It has been widely used in military and civil. Joint detection can cover the shortage of one single detector and gets high detection probability. In this paper, we aim to apply joint detection in optical band. Using active/passive detection means and different optical band, the detecting precision can be improved. First, analyze the characteristics of the targets and background of joint detection system. Here the ORCS and the infrared radiation characteristic of plane is priority research area. From the result of analysis we can see that the passive detecting band 8～12μm optical band is better than 3～5μm optical band.the frame of this optical joint detection system be designed. we make the further studies on the key technology and main influences of every subsystem. Analyzing characteristics of signal and main noise of this system and also studying on the signal processing in this system.We pay more attention to the detection algorithm of optical joint detection system. According to this joint detection system, we applied two algorithms in this system. From the analysis we can see that there is shortage in both two algorithms because of the detection probability is seriously influenced by the change of day and night. In order to reduce the false alarm probability, we adjust the fusion process and bring forward a new algorithm to this system. Finally, from the result of emulate, we can get higher detection probability than only one single detector.更多还原