Filtering and Performance Evaluation for Restoration of Grayscale Image Corrupted by Salt & Pepper Noise Using Low Pass Filtering Schemes

In this paper, different novel low pass filtering schemes are investigated to restore a gray scale image corrupted by Salt & Pepper noise. The filtering methods with different window sizes are applied to corrupted images with varying strengths and different noise probability. We have proposed both subjective and objective methods to evaluate the performance of different filtering techniques for noise removal with an aim to find an efficient filter which will be suitable for real time image restoration applications. Subjective methods have been performed by visually comparing the different restoration characteristics in the output image. The mean square error, signal to noise ratio, peak signal to noise ratio and signal to noise ratio improvement have been used for objective evaluation of the different filtering schemes. Finally, to validate the efficiency of filtering schemes, simulation study has been carried out using MATLAB 5.0.