The images captured by synthetic aperture radar (SAR) are degraded by speckle noise. The features in the SAR images are degraded due to speckle noise and causes misinterpretation of objects. In order to improve the proper interpretation of the objects in SAR image the speckle noise must be seaparated from the image. To remove speckle noise we proposed Deviation controlled soft thresholding in curvelet domain. This technique will improve the limitations of basic thresholding techniques i.e. soft and hard threshold. The results obtained from this approach are superior compared to new wavelet soft thresholding,curveletcyclespin and curveletdenoising techniques.
Volume 11 | 02-Special Issue
Pages: 551-559