A Survey on Techniques Used for De-Speckling of SAR Images

Abstract

Because of the capacity to work at any climate circumstance and infiltration ability through the clouds and soil, SAR (Synthetic Aperture Radar) imagery has been very demanding.SAR imagery system captures images from any satellite or moving object such as aircraft etc. As per the height of a RADAR system increases the resolution of SAR images is also increased. Due to high-resolution images and multiplicative noise known as speckle, it is not easy to analyze these images. So, it has become very important to remove speckle from these images.There are great numbers of de-speckling techniques have been proposed in past. It is very difficult to identify a suitable technique for a de-speckling task. In this article, authors have proposed a literature survey of some techniques.Authors compared best techniques based on their merits and the different parameters used so far for the observation of SAR images. In recent years, the popularity of deep learning has grown tremendously. For de-speckling of SAR images, various improved CNNs and deep learning based approaches have been proposed.