A Denoising Method of Remote Sensing Images Based on Improved BM3D.

Denoising of remote sensing images is always one of the hotspots in the field of digital image processing. The quality of denoising determines the precision of subsequent processing directly. The BM3D algorithm takes the high similarity of image neighborhoods into consideration, which shows good stability and accuracy in the field of image denoising. However, in the case of strong noise, BM3D can't match the image-block precisely, which reduces the denoising effect for the remote sensing image. Therefore, in order to solve the problem of BM3D algorithm during the remote sensing image processing, we propose a local block-matching algorithm based on edge search strategy and residual filtering, according to the high similarity at the edge of the remote sensing image. We compare the proposed algorithm with the classic denoising method, and the experimental results show that the proposed algorithm not only protects the information of edge and detail, but also improves the value of PSNR. A better denoising effect was obtained on remote sensing images.