A Hybrid Algorithm for Speckle Noise Reduction in SAR Images
Nayef Al-Delbah, Kie Bum Eom · International Conference on Aerospace Sciences and Aviation Technology · 2011
Synthetic Aperture Radar (SAR) images are usually impaired by Speckle Noise that appears as granules that result in degrading the fine details in the image. In order to reduce/remove such speckle noise, various processing algorithms have been considered in literature. In this paper, we introduce a Hybrid Algorithm based on spatial domain as well asfrequency domain processing of SAR images so as to allow for better removal of speckle noise and preservation of the fine details. The proposed hybrid algorithm resulted in better than 30dB improvement in the signal to noise ratio; i.e. in significant reduction/removal of most of the speckle noise from the image while preserving of image’s fine details. Theperformance of the implemented hybrid algorithm for speckle noise reduction has been compared with the published respective performance results obtained by other algorithms. Pros and cons of various algorithms are also given along with conclusions and directions for future work.