Fast algorithm for despeckling sentinel-1 SAR data

Speckle noise is an inherent dilemma that alters image processing field particularly synthetic aperture radar images. In order to mitigate its adverse effects, different approaches have been proposed in literature in the last twenty years. However, all approaches suffer from some limits. In particular it is very difficult to find an approach able, at the same time, to preserve image details and provide the solution without requiring high computational complexity and time. This paper aims to test a new despeckling algorithm that is able to jointly guarantee the two previous requirements. The algorithm, based on an evolution of Wiener Filter, modified using Markov Random Fields, is tested and compared on a real Sentinel-1 data. The results are interesting and promising: the proposed algorithm turns to be a useful instrument in case large images or stack of images need to be filtered within a limited time, ensuring good detail preservation.