Comparison of the performance of the newly developed CDWM Filter with Enhanced LEE and Enhanced Frost Filters over the SAR image

Synthetic Aperture Radar (SAR) is an active remote sensing technique. SAR is corrupted by a signal dependent multiplicative noise called speckle noise. Speckle noise limits the data extraction capabilities and degrades the quality of the obtained data. It decreases the potentiality to interpret the image, restricts edge abstraction, image segmentation, target recognition and classification. It is important to carry out despeckling to enable any further usage of SAR data in various possible applications. In this paper, we compare a recently proposed Cornered Difference Weighted Mean (CDWM) filter with two widely use Enhanced Lee and Enhanced Frost filters. The comparison had been done to quantitatively evaluate the performance of each of the three filters using well established evaluation metrics. Based on the results, it is found that the CDWM filter is more computationally efficient and effective in speckle reduction and also preserves fine details in the images. The current study has been done using India´s first C-band (5.35 GHz) SAR based RISAT-1 satellite data.