A versatile denoising method for images contaminated with Gaussian noise
Akshat Jain, Vikrant Bhateja · 2012
Proper choice of denoising filter is a very important requirement for efficient image restoration because most of the filters only reduce the effect of the noise rather than removing it. In this paper, a novel algorithm for filtering of gaussian noise based on the statistics of the robust estimation is proposed. The gaussian noise is replaced by either the computed mean of the adaptively increasing localized window frame or the last processed pixel. Improved Robust Statistics are then applied to obtain the final denoised output. The proposed algorithm is objectively evaluated using Peak Signal to Noise Ratio (PSNR) as figure of merit. Simulation results indicate a marked improvement in the quality of the restored image.