A Study of Images Denoising Based on Two Improved Fractional Integral Marks

In this paper, applying fractional calculus Grumwald-Letnikov definition, a novel image denoising approach based on two improved fractional integral masks is proposed, Two structures of 3×3 fractional integral masks which center is the processed pixel are constructed and discussed. The denoising performance of the proposed fractional integral marks (FIM1 and FIM2) is measured using experiments according to subjective and objective standards of visual perception and SNR values. The simulation results show that SNR of FIM1 and FIM2 is prior to the mean filter and the method in Ref. [8].