Novel effective fuzzy diffusion algorithm for noise removal

Anisotropic diffusion is widely used for noise reduction. The performance of anisotropic diffusion, in general, depends on the shape of the energy surface. The partial differential equation model is established and analyzed in the continuous domain while is implemented in the discrete domain. Therefore, the anisotropic diffusion bears some fuzziness due to the approximation. We present a novel noise removal algorithm based on fuzzy logic and anisotropic diffusion theory. The experimental results demonstrate that the proposed method has the advantage of maximizing noise reduction and preserving fine details of the images. In addition, the method can enhance the contrast of the images well.