De-Speckling of Synthetic Aperture Radar Using Discrete Fourier Transform
Sanjay S. Shitole, Vijal Jain, V. S. K. Vanama · 2020
The role of Synthetic Aperture Radar (SAR) images is inevitable in remote sensing applications. One of the major concern in SAR imagery is that basic textures are generally affected by multiplicative speckle noise. Speckle noise is a consequence of image formation under coherent radiation, but it often carries useful information about the scene being imaged. However, speckle noise is considered undesirable as it damages the resolution and affects the tasks of human interpretation. The proposed framework uses Boxcar filter for reduction of speckle-noise whilst retaining the point targets, edges and texture with the inclusion of Discrete Fourier Transform (DFT) in the speckle reduction framework. A novel technique is propounded, which aims at having a fine control on the high-frequency components by tuning the radius of the frequency rectangle.