An Improved Switching Median filter for Uniformly Distributed Impulse Noise Removal
Rajoo Pandey structure. Another approach followed in [8] uses a differencetype noise detector and the noise detection-based adaptive medium filter. The boundary discriminative noise detection (BDND) filtering scheme proposed in [9] detects the impulse noise by employing two different size of filtering windows before the filtering operation. Most of these filtering techniques assume the presence of salt and pepper type of impulse noise. The detection of salt and pepper type of noise is relatively easy as there are only two intensity levels in the noisy pixels. However, the study reveals that in case of uniformly distributed impulse noise, these techniques do not perform well. In this paper a new algorithm is presented which improves the performance of switching median filter as a result of efficient detection of impulse noise when the impulse amplitude is uniformly distributed. The paper is organized as follows. Section II discusses the impulse noise removal technique using switching median filters. Section III presents the proposed noise detection algorithm for uniformly distributed impulses. The simulation results with different images are presented in section IV to demonstrate the efficacy of the proposed algorithm. Finally, conclusions are given in section V. II. IMPULSE NOISE REMOVAL The impulse detection is based on the assumption that a noise pixel takes a gray value which is substantially different than the neighboring pixels in the filtering window, whereas noise-free regions in the image have locally smoothly varying gray levels separated by edges. In the switching median filter, the difference of the median value of pixels in the filtering window and the current pixel value is compared with a threshold to decide about the presence of the impulse. We assume that the image is of size M×N having 8-bit gray