Optimization of Grunwald-Letnikov's (G-L) based Fractional Filter Used for Image Enhancement

This Paper aims at finding out the optimized value of the fraction used in G-L based fractional filter for image enhancement techniques in digital image processing. To start with, a comparative study of fractional order filter and a integer order kernel filters like Sobel filter are made by performing filter action on MRI images. The performance of the filters are analyzed by adding random Gaussian noise. Evaluation of the performance is done in terms of PSNR and MSE and thereby compared. From the analysis the optimized value of the fraction is determined.