SAR Image De-noising Based on Nuclear Norm Minimization Fusion Algorithm

Synthetic aperture radar (SAR) images play a quite important role in military and environmental monitoring. But the SAR image was greatly affected by coherent noise, which affects its application in the subsequent image analysis. In most of the SAR image de-noising algorithms in hand, the same operation is applied to the whole SAR image, which leads to artificial texture or edge blur. In order to overcome this shortcoming, this paper proposed a new SAR image de-noising method based on nuclear norm minimization (NNM) fusion algorithm. The noisy SAR image is de-noised by two different algorithms, and two de-noising images are fused to final de-noising image based on nuclear norm minimization fusion algorithm. Experimental results show that the proposed algorithm not only effectively improves the visual effect and objective indicators of de-noising image but preserves the local structure of the image better.