Remote Sensing Image Enhancement via Robust Guided Filtering

Abstract

Remote sensing image enhancement methods have to preserve the original reflectance values as possible as they can, whereas emphasising the edges and increasing the contrast. In this study, a remote sensing image enhancement method based on robust guided filtering is proposed. We propose a multiscale decomposition with the robust guided filtering to obtain the approximation and detail layers of the image. Then the extracted details are amplified and added to the approximation layer to obtain the enhanced image. Both visual and quantitative comparisons show that the proposed method has a better preservation capability than the former methods, as well as a better contrast improvement along-with edge enhancement.