Image De-noising Based on Nonlocal Diffusion Tensor

The nonlocal structure tensor of images is defined by using the nonlocal spatial gradients. The eigenvectors of the nonlocal structure tensor consist in a characteristic space for the image, based on which the nonlocal diffusion tensor is constructed. Utilizing the nonlocal diffusion tensor, we introduce the nonlocal anisotropic diffusion model for image de-noising. The model we proposed differs from the local anisotropic diffusion in that, not only neighboring pixels but also pixels faraway with similar intensities are concerned in our model. The main advantage of the model is that it protects textures much better than the local model.