Grey Theory Based Evaluation for SAR Image Denoising

To evaluate the quality of denoised SAR images, this paper discusses some performance parameters, and proposes a grey-theory-based method. The main idea of the method is to extract some comparative sequences and a reference sequence, respectively standing for quality of the denoised SAR images and the noise-free image, after all the denoised images involved are concerned. And then, an overall performance index is acquired from the grey relational degrees between the two kinds of sequences. The feature of the method is that it not only does not require any real images, but also is inexpensive in processing time. Pilot experiments show that the method is effective and precise. Potential application includes automatic selection of the optimal algorithm and intellectual comparison among SAR denoising algorithms.