An Efficient Approach for Salt and Pepper Noise Minimization using Multi Stage Thresholding and Image Restoration
Vijaya R Arjunan, Kishore Bhamidipati · 2013
In this paper, a modified two phase Tolerance based Arithmetic Mean filtering (TSAMFT) is verified for noise minimization in images which are heavily contaminated by Salt and Pepper noise (Intensity spikes). Two additional features are implemented in the conventional Arithmetic Mean approach. In the first phase, Arithmetic Mean is calculated only with unaffected pixels and in the second phase, a deduced Tolerance value is used for replacement of pixels. The proposed method is generalized both to grey scale and color images with varying scale measure. Experimental results show that, the proposed algorithm has definite advantages over non-linear filtering techniques in terms of edge preservation and image quality. Quantitatively, the proposed approach is competent with other two stage noise removal methods.