混浊介质中的目标识别及其在交通安全中的应用  

Object recognition in turbid media and its application in traffic safety

作　　者：徐新苗[1] 徐兰青[2] 郑勇平[2] 林宏翔[2] 

机构地区：[1]福州大学机械工程学院,福州350002 [2]福建师范大学,福州350007

出　　处：《激光杂志》2008年第6期53-54,共2页Laser Journal

摘　　要：结合迷雾天气中车辆交通安全研究的核心内容——目标识别问题,利用斯托克斯/米勒表述和蒙特卡罗方法探讨偏振光在大气这类混浊介质中的传输规律,利用反射米勒矩阵对混浊介质中隐藏的目标物体进行识别。数值计算结果表明,在一定距离之内,利用反射米勒矩阵计算的解偏度能够较好地分辨出混浊介质中隐藏的不同物体。结合迷雾天气的具体特点,分析有效识别距离的大小,定量说明了识别距离大于能见距离,高于直接利用人眼识别的水平。Object recognition in a foggy weather has long been a tough trouble in traffic safety study. In this paper Stokes/Mueller formula was introduced to investigate polarized light transportation in turbid media such as atmosphere; Monte Carlo method was used to calculate reflected Mueller matrix and corresponding degree of depolarization （DOP） from a turbid media containing hidden object. Results showed that the 2 - Dimension distribution of DOP can well discriminate different objects within limited distance. For a foggy weather the calculated effective recognition distance could be several times of visibility distance.

关 键 词：大气光散射 目标识别 交通安全 斯托克斯/米勒表述 蒙特卡罗模拟 

分 类 号：O439[机械工程—光学工程]