Image de-noising by dual threshold median filtering for random valued impulse noise

In the field of digital image processing^[4], noise removal is always a critical process. In this paper we proposed an enhanced method of image de-noising. The purpose of this new method is to improve the signal to noise ratio (SNR) of de-noised image and get more better image, especially when image corrupted by high noise density. We improved the median filter algorithm, and get comparatively better results than previous methods. The mathematical analysis shows that this process improve the PSNR^[8] (Peak signal to noise ratio) at high density noise level. This method used the combination of median and average functions. Two levels of threshold and improved median value remove the noise much effectively. The mathematical analysis shows that the analysis of the noisy pixels and use of noise-free pixels for de-noising purpose produces much better results and gives better quality de-noised version.