基于被动毫米波图像的金属目标特征提取方法  

Feature Extraction of Metal Targets Based on Millimeter Wave Radiation Image

作　　者：刘静[1] 张光锋[1] 褚庆国 郑振忠[2] 陈庆昌[2] 周璐艳 

机构地区：[1]南京理工大学电子工程与光电技术学院,南京210094 [2]北京特种车辆研究所,北京100072

出　　处：《计算机测量与控制》2017年第5期169-172,共4页Computer Measurement &Control

基　　金：国家自然科学基金项目(61371038);江苏省高校优秀中青年教师和校长境外研修计划

摘　　要：针对探测金属目标时其中心难以确定的问题,提出了分割被动毫米波辐射图像过渡带的算法以精确获取其几何尺寸及特征;通过实测金属立体目标的辐射数据矩阵进行成像,对毫米波辐射图像进行基于Canny算子的边缘检测与分析,结合金属目标辐射特性以及天线参数分割目标边缘过渡带,计算金属目标的面积、周长并提取其中心位置;实验结果表明,该方法误差控制在9.3%以内,可以精确提取金属目标几何特征及目标中心。To solve the problem that the accurate center of metal targets in detection is difficult to extract,a method about reserving and dividing the transition zones in passive millimeter wave(PMMW)radiometric images was presented to extract the characteristics of metal targets.The radiometric data matrix of three-dimensional metal target was collected and imaged.Canny edge detector was used to detect the edge of the image and analyze the characteristics.Based on the characteristics,the perimeter and area can be calculated as well as the center.The experimental results show that the method can detect the characteristics of metal targets efficiently and exactly with error in 9.3%.

关 键 词：被动毫米波 辐射图像 金属目标 边缘提取 中心提取 

分 类 号：TN911[电子电信—通信与信息系统]