Parameter analysis of Douglass-Gun ADI scheme on anisotropic diffusion model for image denoising

• Published in: AIP Conference Proceedings
• Main Author: Halim S.A.; Aziz N.A.A.
• Format: Conference paper
• Language: English
• Published: American Institute of Physics 2024
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85194197958&doi=10.1063%2f5.0208857&partnerID=40&md5=a6265d5f50c6cb067fc759d4b196e4fb

Anisotropic Diffusion (AD) is a model that is used for edge preservation in image denoising process. The modification of the AD model is developed to overcome the problem of blurring effect in image. In order to maintain the edge structure of the denoised image, the optimal estimation of the parameters in the AD model is the main contributor to remove the noise smoothly without over-smoothing effect that may destroy the small structures of the images. In this paper, Douglas-Gunn Alternating Directional Implicit (DG-ADI) scheme is implemented into the AD model to remove the noise that corrupted the images since the scheme is unconditional stable and able to solve the nonlinear equation with multidimensional spaces efficiently. The parameter analysis on the gradient threshold, K parameter selection and its performance are evaluated based on different coefficient functions. The performance of the scheme is tested on several grayscale images and the implementation of the algorithm is using MATLAB R2013a. Experimental results show that the DG-ADI scheme able to remove noise with high range parameter and small-time step based on the Peak Signal to Noise Ratio (PSNR), Mean Structural Similarity Index (MSSIM) and Universal Image Quality Index (UIQI). The used of DG-ADI scheme in solving the AD model is useful in removing noise well while preserving the fine structures of image. © 2024 Author(s).