A hybrid image denoising method based on clustering and PDE

This paper provides an efficient method based on the combination of hierarchical clustering along with the capability of PDE, FFT and color domination. Then for the edge point selection decomposition has been performed. It is applied with the clustering mechanism so that data points are separated. Then by similarity ranking alike data points are separated and decomposed. By this process, noise can be separated and other image proprieties along with the alikeness are separated. The color domination, PDE and FFT combination have been applied. This is applied on the data obtained from the previous process. This step provides the color based separation and error filtration. PSNR values have been used for the comparative study. The obtained results have higher PSNR then the previous approaches shows the effectiveness of our approach.