Vectorial total variation model for multi-channel SAR image denoising
多通道SAR图像滤波的向量总变分模型

This paper proposed an unsupervised technique for detecting changed areas between multitemporal SAR images. Different with the original ones, the clustering method was used here to find the change map by minimizing mean square error with evolution algorithm. After introducing the image character, a new search strategy in Memetic algorithm was given here, which adjusted the local search algorithm according to the current detection result. The approach was distribution free and did not need priori knowledge. The experimental results obtained on the real SAR images showed that the proposed method had a higher convergence speed than GA,ICSA and original MA, the detection results demonstrated the effectiveness of the proposed algorithm.