基于CCA差异图融合的SAR图像变化检测算法

SAR Image Change Detection Algorithm Based on CCA Difference Graph Fusion

针对SAR图像差异图融合方法操作繁琐、参数不能自适应选取、计算复杂这一问题,提出了基于典型相关分析(CCA)差异图融合的SAR图像变化检测算法。该算法首先构造比值图、对数比值图,然后经过中值滤波,将滤波后的比值图和对数比值图转化为列向量并将两组列向量进行线性组合,在典型相关极大的条件约束下,得到最佳的线性组合,再将此最佳线性组合的列向量转化为原差异图大小的矩阵,即可得到融合后的差异图。对融合后的差异图应用FCM聚类得到聚类效果图,而后对聚类效果图应用Growcut分割算法处理得到最终的变化检测结果。由实验仿真结果可以得出所提基于CCA差异图融合的SAR图像变化检测算法具有很高的变化检测精度,是适用可行的。

Aiming at the problem that the SAR image difference graph fusion method is cumbersome and the parameters can not be selected adaptively,this paper proposes a SAR image change detection algorithm based on typical correlation analysis difference graph fusion.The algorithm first constructs the ratio graph and the logarithmic ratio graph,and then transforms the filtered ratio and logarithmic ratio graph into the column vector and then linearly combines the two sets of column vectors.After the typical correlation condition Under the constraint,the best linear combination is obtained,and then the best linear combination of the column vector into the original difference in the size of the matrix,you can get the difference after the convergence map.The FCM clustering is used to get the initial change test result after the fusion difference map,and then the final change test result is processed by the Growcut algorithm.The experimental results show that the proposed algorithm has a high accuracy of detection and is feasible.