车载UWB SAR图像预筛选中CFAR检测的适用性分析  

Applicability Analysis of CFAR Detection in Prescreening the UWB SAR Image of Vehicle-mounted System

作　　者：杨延光[1] 初宁[1] 周智敏[1] 王玉明[1] 宋千[1] 

机构地区：[1]国防科技大学电子科学与工程学院,长沙410073

出　　处：《武汉理工大学学报（交通科学与工程版）》2010年第5期864-868,共5页Journal of Wuhan University of Technology（Transportation Science & Engineering）

基　　金：国防重点预研项目资助

摘　　要：车载前视探测系统通常在行进间探测大面积复杂区域,对图像预筛选算法的实时性和环境适应性提出了更高要求.文中分析CFAR检测器用于车载UWBSAR图像预筛选的可行性,提出一种基于循环迭代的统计量快速计算方法,以满足工程应用中对实时性的需求;给出实际应用中选择CFAR检测器参数的准则和方法.实测数据的试验结果表明:在给定窗尺寸下,本文提出的方法可使得CFAR检测器的效率提高2.21倍,所给出的参数确定方法是合理可行的.Vehicle-mounted forward-looking detection system usually explores the large and complicated area during move along.Thus,it demands that the prescreening algorithms applied to SAR image should be robust and real-time.Firstly,we analyze the applicability of the constant false alarm rate （CFAR）detector to the prescreening of vehicle-mounted ultra-wideband synthetic aperture radar （UWB SAR）images.Then,a fast computing method of statistic based on circularly iterative operation is presented to meet the needs of real-time processing.Finally,we give the rules and methods to select the parameters of CFAR detector.Experimental results of real data show that the proposed method can improve the efficiency of CFAR detector 2.21times under the given window size,and the suggested approaches for parameters determination are reasonable and feasible.

关 键 词：合成孔径雷达 图像预筛选 恒虚警率 循环迭代操作 

分 类 号：U495[交通运输工程—交通运输规划与管理] TN957.51[交通运输工程—道路与铁道工程]