论文部分内容阅读
为了克服传统的基于合成孔径雷达(SAR)图像局部统计特性的降斑算法的缺点,提出了基于区域分类、自适应滑动窗和结构检测的联合降斑算法.首先,联合降斑算法对当前区域进行区域分类,并直接保留强边缘结构和点目标.接着,联合降斑算法对均匀区域和弱边缘结构区域进行滑动窗的自适应增长,从而获得合适的滤波窗口.最后,联合降斑算法对新的滤波窗口使用区域分类.如果滤波窗口属于均匀区域,则直接使用均值滤波;如果滤波窗口属于边缘结构区域,则进一步使用结构检测,并且选择窗口内的均匀子区域作为最终的滤波区域.降斑实验表明,联合降斑算法可以有效滤除均匀区域和边缘区域的斑点,同时对强边缘结构信息和点目标有较好的保留作用.
In order to overcome the shortcomings of the traditional speckle reduction algorithm based on local statistic characteristics of Synthetic Aperture Radar (SAR) images, a joint speckle reduction algorithm based on region classification, adaptive sliding window and structure detection is proposed.First, And then keep the strong edge structure and the point target directly.And then, the joint speckle reduction algorithm is used to adaptively increase the sliding window in the uniform region and the weak edge region to get the proper filtering window.Finally, If the filter window belongs to the uniform area, then the average filter is used directly. If the filter window belongs to the edge structure area, the structure detection is further used and the uniform sub-area in the window is selected as the final filtering area. Speckle experiments show that the joint speckle reduction algorithm can effectively filter out the spots in the uniform area and the edge area, and retain the information of the strong edge and the point target well.