一种基于双阈值区域分割的SAR图像目标提取方法  

Algorithm for SAR image target detection based on double threshold segmentation

作　　者：刘开刚[1] 许梅生[1] 李维 

机构地区：[1]炮兵学院,合肥230031 [2]襄樊陆军预备役炮兵旅,襄樊441104

出　　处：《国外电子测量技术》2008年第3期3-6,共4页Foreign Electronic Measurement Technology

摘　　要：本文提出一种基于"双阈值"的区域生长算法。该方法首先对图像中感兴趣的区域进行分割,通过阈值的选取对已分割区域进行合并与剔除虚警来提取感兴趣的目标。采用国产机载SAR图像,以大中型建筑物为例进行了实验。实验结果表明该方法能够有效地提取目标特征,进行区域分割。A double threshold segmentation algorithm is presented in the paper. The algorithm segment regions of interest in the image firstly. Targets of interest in the regions of interest which have been incorporated and picked-out false alama are then detected by selecting threshold. This paper is based on large or medium buildings in SAR image which are achieved by homemade airborne synthetic aperture radar. Experiments show that the method can detect targets effectivelv.

关 键 词：SAR图像 区域分割 目标识别 

分 类 号：TN911.73[电子电信—通信与信息系统]