An adaptive diffusion coefficient scheme for image denoising based on anisotropic diffusion

For image denoising and edge keeping, an adaptive diffusion coefficient scheme based on the traditional P-M model is proposed in this paper. The edge detection operator which reflects the edge information is introduced to obtain the optimum gradient threshold, not as same as using the constant value K in the classical diffusion coefficient equation. Experimental results either from the ideal images or form the practical two-photon microscopic images have shown that the proposed approach for image denoising is better than the fixed value K in the diffusion coefficient function both at edge preserving and noise removing.