Adaptive Weighted Least Square Filter for SAR Image Denoising and Edge Preservation

Synthetic Aperture Radar (SAR) image processing is a technique capable of producing high-resolution images even under adverse weather conditions. However, SAR images are often affected by speckle noise, which degrades edge detection and image details. This study presents the development of a filter for reducing noise in SAR images by utilizing the Hessian matrix to analyze internal image structures and adjust the filter's weighted factors accordingly. The proposed filter was tested on both synthetic and real SAR images, demonstrating superior noise reduction performance compared to widely used adaptive speckle noise reduction filters. The results indicate that the developed filter effectively preserves image edges and important details.