An efficient image denoising method based on KPDE

In this paper a k-means based PDE has been applied for image denoising. In this approach first data pre-processing mechanism has been applied. The next procedure is for the image denoising. In this process the pre-processed image has been selected. Gaussian noise has been added in terms of noise percentage. Then object based clustering and decomposition has been applied for efficient data point selection. For this k-means algorithm has been applied. By this process object point cluster has been obtain. The main benefit by this approach is it is able in finding the decomposition as well as the similar point by the similarity ranking and matching. PDE-FFT hybridization has then been applied on the clustered data for the final noise separation. Then the PSNR values have been calculated for the comparative study. The results indicated that our approach has the capability in better noise removal in terms of previous method.