Impulse Noise Detection-based Image Denoising

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.For image noise removing, the traditional linear filter has some inherent shortcomings. Therefore a new class of nonlinear filters structure has been developed in recent years. Due to its special performance in image noise removing, it has become a research focus of image noise removing .In this paper, we have done some further studies on traditional non-linear filter. As a result, a.new filter algorithm has been proposed.The main tasks of this paper are as following: we study the traditional nonlinear filter-median filter carefully, based on the traditional median filter, the improved median filter, the weighted median filter, the min-max median filter, the switch median filter, the adaptive median filter and a new weighted adaptive median filter, we know theirs advantages and disadvantages. We change a little, a new median filter has been proposed and then its main proper have been analyzed. We can find that it is better on removing the noise.