Image Denoising using Variation Perona-Malik Model Based on the Variable Exponent
Image enhancement can be achieve better performance by using partial differential equation (i.e.) Perona-Malik model and Heat equation than using different conventional filters. However traditional Perona-Malik model gives some artifacts and introduce staircase effect on the resultant image. The proposed method variation of Perona-Malik Model introduces edge indicator as they preserve the edges and smoothens the image. As edge indicator uses constant and variable exponent which gives different conclusion of the resultant image as efficient method. For better enhancement the pre-processing Gaussian filter and Wiener filter are used before diffusion process as they removes the noise completely on the image. The threshold parameter and smoothing controller of diffusion process are varied for further analysis of the performance. The proposed method using pre-processing as they smoothens the image and the boundaries of the image are enhanced. The performance of the resultant images are measured using quality metrics and tells how efficient works on the proposed method.
