SAR image filtering algorithm based on adaptive hexagonal window
Ming Bo Zhu, Bailong Yang · 2018
In view of the contradiction between the traditional Lee filtering algorithm in suppressing the speckle and preserving the details of the edge, a new adaptive window is proposed to improve the algorithm of Lee filtering.The filtering window uses a hexagonal design. In the filtering process, each pixel of the image is first traversed through the window, and the pixel is re-assigned in the window with a certain weight. Then, on the newly generated image, combined with the Lee filtering algorithm, according to the pixel uniformity in the window with the threshold size to adaptively decide whether to expand the window, use a large window in a uniform area, non-uniform area selection of small windows, and then the image filtering. Experiments show that the proposed method can avoid the speckle noise and prevent the detail information from being filtered out, and the filtering effect is better than the general window, which has certain engineering application value.