Research On Parameter Estimation Of Partial Difference Equation Model In The Optical Image Denoising

Image denoising is one of the most important steps in image processing. Its purpose is to enhance the SNR between original image and denoised image, improve the character of image, and reduce the influence to the following image processing as far as possible. The traditional methods of image denoising always filter the high-frequency component of image. But the image edges also distribute in high-frequency component. So they are always blurred while filtering the noise in image.Partial differential equation filters are well-known for good processing results, but we must first solve an important question on optimal parameters selection because it affects the denoising results and the stability of the equation directly, which include the selection of diffusivity and optimal stopping time. In this paper, I first analysed the enduce, principle and its development, based on which, I pointed two important questions of the PDE denoising model: selecting model and selection of parameters.The selection of optimal stopping time for PDE diffusion filtering was studied. This paper analyzed the optimal stopping time criterion based on SNR, and expatiated different image noise model and noise levels on the impact of experimental results, comparing with the minimum correlation coefficient criterion and through experiments. And on the basis of the experiments, we pre-processed the image by introducing the anisotropic Gaussian filtering, Simulation results showed that the SNR criterion can approach the ideal stopping time earlier and better than the minimum correlation coefficient criterion.At last, I construe the basic principle, advantages and disadvantages of the P-M model. A method of choosing parameters is given at the end of this paper.