Discussion Of The Image Filtering Method And Realization Of MATLAB

Image restoration or filtering technology has 40-year history. Image restoration modern technology has a broad range of applications. In the process of image imaging, copy, scan, transmit, display and so on, it is inevitable that image quality has been lowered, such as blurred images, noise interference. But in many application fields, they need clear, high-quality images. So image restoration (such as removing noise, defuzzifier) is very significant. Image restoration is processing the lower image quality, and improving the image quality, as well as restored to the ideal image with no degradation. It is the basis of image processing, pattern recognition and machine vision, therefore it is widely concerned and used in astronomy and remote sensing imaging, medical images. In the field of image processing, image restoration is the most important and basic one of the research, and it has important value of theory and practice.This article firstly describes the concept of image restoration or image filter, and then discusses the mathematical model about image degradation and image noise. Two main problems of image restoration are to remove noise and to defuzzy. To sum up, the traditional method of image restoration, can be divided into inverse filtering method of algebra and spatial domain filtering method. Inverse filtering method has classical inverse filtering, Wiener filtering, Kalman filtering, etc. Algebraic methods can be divided into pseudo-inverse method, singular value decomposition pseudo-inverse method, Wiener estimating method and bound image restoration, etc. This article discusses the classical inverse filtering, Wiener filter, and bound least-squares image restoration method. As most information of the image found in the part of the edge, therefore Image filtering should be asked to remove the noise and fuzzy images, while maintaining the image details. Median filter is a non-linear image processing technology based on permutation and statistical theory that can effectively suppress noise. It has the advantages of simple and fast operation median filter can not only remove noise but also protect the edge of the image, in order to achieve more satisfactory restoration effects. But it should not be used for some image that has more details, especially more details of point, line, and the steeple of the image. In this paper, the median filter of optimization and improvement should be used to solve these problems, in order to get high-quality image restoration.