A Novel Image Denoising Algorithm based on GA, PDE and TV

Magnetic Resonance Imaging (MRI) images are frequently corrupted by Rician noise during image gaining or transmission, which makes them worse than original. With the object to remove noise and lessen the loss of details as low as possible at the same time, we proposed a novel image denoising algorithm based on Genetic Algorithm (GA), Partial Differential Equations (PDE) and Total Variation (TV). First, the TV and the 4th order PDE of kth image will be calculated and weight coefficients will be given. Second, by adjusting weight coefficients based on GA algorithm, it will find the optimal standard deviation of kth image. At last, PSNR of kth image and k+1th image will be compared to find whether the algorithm is over. Experimental results show that our proposed algorithm is more effective in removing Rician noise compared to some other traditional algorithms.