Noise Removal Through the Exploration of Subjective and Apparent Denoised Patches Using Discrete Wavelet Transform

ABSTRACT

The effective elimination of noise from the original image has emerged as a hard nut to crack for the researchers. The hunt for effective image denoising techniques continues to be a strenuous task at the cross junction of the functional investigation and statistics. Though there are many approaches to image denoising, the performances of them are comparatively low. In this paper, the internal and external data cubes are smartly configured so as to arrive at the identical patches from the respective noise-contaminated and web images. In this regard, there are two phases employing diverse filtering methods intended for the reduction of noise. In the initial phase, a graph-based optimization technique is introduced to effectively enhance the patch harmonizing in the external denoising. In the second stage, the noise is significantly scaled down by removing the respective internal and external cubes. Discrete Wavelet Transform (DWT) filtering approach is utilized for achieving the superlative precision in denoising image in relation to the existing filters.

KEYWORDS:

• Correlations
• Discrete wavelet transforms
• Filtering
• Image restoration
• Optimization

Additional information

Notes on contributors

K. Sakthidasan alias Sankaran

K Sakthidasan alias Sankaran received the bachelor’s degree in electronics and communication engineering from Anna University and master's degree in embedded systems technology from SRM University. He has obtained his doctorate degree from Anna University. He is working as an assistant professor in ECE, Adhiparasakthi Engineering College, Melmaruvathur. He has published many papers in reputed journals and international conferences. He has also published three books in his career.

V. Nagarajan

V Nagarajan is the professor at APEC, Anna University, Chennai, India. His MTech and PhD degree are from Pondicherry University, India. His research interest includes wireless communications and image processing. Email: [email protected]