Median Filtering Image Denoising Algorithm Based On Noise Pixel Detection

During the process of formation and transmission, images are often disturbed by different types of noises, which degrade image quality. Various linear or nonlinear filtering methods are employed to reduce or remove noises. Because of better performance of eliminating impulse noise, the median filtering method is used early to remove impulse noise. However, the simple median filtering algorithms have an impact in all pixels in noisy images and often incur over-smoothness of image's details. Thus, it is important to improve the simple median filtering algorithms such that image's details are preserved. This paper mainly investigates the median filtering methods based on the noise pixel detection. This paper includes mainly two parts as follows. In the first part, a spatially adaptive fuzzy filter is presented for the restoration of images corrupted by impulse noise based on local spatial structure information. The experimental results show that the proposed method can not only remove impulse effectively but also preserve image's details well. In the second part, based on the TNOM algorithm, we give its general structure and propose some improvements.