Denoising of images using Thresholding Based on Wavelet Transform Technique
Gurjinder Kaur, Meenu Garg, Sheifali Gupta, Rupesh Gupta · IOP Conference Series Materials Science and Engineering · 2021
Abstract This paper presents an efficient method based on thresholding for images that are corrupted due to Gaussian noise. In order to achieve this task, wavelet transforms and various thresholding techniques have been applied. But the selection of the thresholding technique is restricted due to their widespread use in image denoising application. In this paper, a more efficient thresholding scheme named as neigh shrink sure is studied by incorporating the neighboring wavelet coefficients. Different thresholding techniques like Bayes shrink and Neigh shrink sure algorithms are applied to different images. The results are obtained using 3 different wavelets db-4, sym-4, and coif-4. It has been observed that the use of coif-4 transform produces better results as compared to other wavelet transforms. Further, it has been analyzed that Neigh Shrink sure method performs better in de-noising of corrupted images in comparison to Bayes shrink thresholding technique by possessing higher Peak Signal-to-Noise Ratio (PSNR) and lower Mean Square Error (MSE).