A fast efficient restoration algorithm for high-noise image filtering with adaptive approach

The video signal is easily interfered by noise over the wireless system. The well-known "median filter" and its modified methods could completely remove the impulse noise when the ratio of noise/image is low. However, if the noise ratio becomes high, the filtering performance degrades seriously. In this study, an adaptive algorithm is proposed to reduce the impulse noise for high noise images. To avoid filtering image edges, the noise reduction ability is adaptive according to the image content. The main techniques include the adaptive noise detection and the non-linear low-pass filter. Experimental results show that the proposed method is capable to provide a better picture quality than the median filters and which is faster than alternatives. (C) 2004 Elsevier Inc. All rights reserved.