SAR图像组合分割算法  

Compounding Segmentation for Synthetic Aperture Radar Images

作　　者：刘振华[1] 毛士艺[1] 袁运能[1] 

机构地区：[1]北京航空航天大学电子工程系,北京100083

出　　处：《电子学报》2003年第6期833-836,共4页Acta Electronica Sinica

基　　金：国家自然科学基金 (No 69831 0 4 0 )

摘　　要：由于存在乘性斑点噪声的影响 ,给SAR图像分割造成很大困难 ,本文研究利用最大似然区域增长分割算法 ,提出区域增长前引入自适应边缘检测 ,图像分割性能获得较大提高 .介绍了我们研究的分割性能优异的模拟退火分割算法 ,提出了分割性能接近模拟退火并且在SAR图像分割中比较实用的模拟退火和最大似然混合分割算法 ,并给出了这些分割算法的分割性能比较 ,以及在实际分割SAR图像时应当如何选择分割算法 .The multiplicative nature of the speckle noise in SAR images has been a big problem in SAR image segmentation. The maximum likelihood region merging segmentation algorithm is studied and employed. An adaptive edge detect procedure is introduced into the algorithm here and better segmentation performance of the improved algorithm with the adaptive edge detect is involved. A simulated annealing segmentation algorithm with excellent performance is studied and introduced here, and a hybrid algorithm which is practical in SAR image segmentation combines simulated annealing algorithm and maximum likelihood algorithm together to achieve a similar performance with the simulated annealing algorithm is also proposed. Segmentation comparison between these algorithms is given as well as instructions of how to select the segmentation algorithms in practice.

关 键 词：合成孔径雷达 分割 自适应 模拟退火 最大似然 

分 类 号：TN957[电子电信—信号与信息处理]