Digital Image Noise Cancellation Algorithm Research

Digital image processing is a new technology which comes out in 1960s.With the development of the computer technology, digital image processing changes very quickly. Image noise removing is an important branch of digital image processing. Image de-noising is one of the widely discussed topics in this field. It is inevitable to avoid image affected by noise pollution that existed in the process that the environment,equipment and transmission, the image acquisition in the reality. So how to reduce the noise is the important step to image preprocess. It is a very important pretreatment steps to minimize noise by an appropriate methods in further edge detection, image segmentation, feature extraction, pattern recognition.This paper's Contents and the framework are as the following:(1)The first chapter introduces the background, the purpose and significance of the image de-noising. And also points out the current problem and difficult.(2)The second chapter we discuss image de-noising on the theoretical foundation of the concept of image noise, noise sources and characteristics of the image and the quality of the evaluation criteria.(3)The three chapter we introduce the median filtering method of principle, character, and de-noising performance improvement in the value of filtering methods, in the extreme value in the value of median filtering, the weight median filter and improved the median filter, as well as better results in the value of the adaptive filter algorithm.(4)The fourth chapter introduces the wavelet transform and partial differential equations in the application of de-noising, especially wavelet transform in the de-noising focused on the application, on the course of development of wavelet transform and wavelet image de-noising threshold method and analysis its existing deficiencies and improving methods.(5)The fifth chapter proposed a new method based on the direction of image de-noising, the method from eight different directions to judge noise, we use a new the adaptive way to determine the threshold value, that makes it have a good effect. About noisy pixel, we use a method for adaptive identification, so that each pixel has an appropriate threshold standards, we use the penalty function to curb and reduce non-point misjudgment of noise and image details fuzzy adaptive threshold identified. The experimental results show that this new algorithm has a better image de-noising protection and performance details(6)The sixth chapter introduces the image de-noising trend of development, and also sets forth this next step will be to solve the problems.This paper's Achievements and explained are as follows: (1)In this paper, we introduced some usually used de-noising methods, and summarizing some of the advantages and disadvantages of these methods, and introduces a number of improve algorithms in the spatial domain and transform domain in the paper.(2)In this paper, we proposed a new method based on the direction of the noise removal algorithm, simulation results prove that this method can also achieve good results. But what is needed to say is that noise used in the experiment is ideal, and it always adding Gaussian noise or Salt & Pepper. But in fact, noise is more complex. So many de-noising methods should be used when we want to get an ideal performance.