Denoising by Anisotropic Diffusion in ICA Subspace

In this paper, we propose an image denoising method that incorporates anisotropic diffusion and independent component analysis (ICA) techniques. An image is decomposed into independent component coefficients, and adaptive anisotropic diffusion is applied to these coefficients. The number of diffusion iteration is determined by the intrinsic properties of the components. The proposed method achieved much better noise suppression compared with other well-known denoising approaches, particularly in denoising of images with extremely low signal-to-noise ratio (SNR). The effectiveness of the method is demonstrated by experiments on x-ray images and electron micrographs.