Application Of Anisotropic Diffusion Equation In Image Processing

Image processing and analysis is a major field of information science and engineering, whose main projects are image enhancing, image restoration,image segmentation and edge detection. During the forming, transferring and memorizing of the digital images, the images are often corrupted by different kinds of noises because of the defection of the imaging system, transferring medium and memorizing equipment. Therefore, the quality of image became worse. Image de-noising is a key point in the fields of image processing. Traditional method of image de-noising is to filter high frequency of image. It is well known that some detail information(edge etc.) are included in the field of high frequency, so edge is blurred while de-noising.Recently, a new method based PDE theory emerge in the field of image processing and attract attention of more and more people. The method came through the process of linear to nonlinear and isotropic diffusion to anisotropic diffusion. This paper present leading models simply, and study P-M nonlinear diffusion model and J.Weickert's tensor diffusion model in detail. In order to avoid disadvantages, I bring forward a new method based J.Weickert's tensor diffusion model. Result of simulation experiment showed that we got a better image after processing by the mew method.