Efficient Color Image Denoising using DWT-based Noise Estimation and Adaptive Wiener Filter
Part Of
Proceedings - 2024 8th International Conference on Imaging, Signal Processing and Communications, ICISPC 2024
Start Page
47
End Page
51
ISBN (of the container)
979-835036715-7
Date Issued
2024-07-19
Author(s)
Yuan-Kang Lee
Abstract
The majority of the existing denoising algorithms either suffer from suboptimal denoising performance, making them prone to failure in specific scenarios, or demand significant computational resources, thereby substantially reducing their practical applicability. For real-world noisy color images, the noise level varies across the R, G, and B channels. Hence, the paramount challenge of real-world color image denoising is how to achieve the most accurate image noise estimation in different channels and execute denoising with efficiency. Our proposed image denoising method integrates an improved Discrete Wavelet Transform (DWT)-based image noise estimation method with the enhanced adaptive Wiener filter. Experimental results shown that our proposed algorithm not only maintains excellent denoising performance but also exhibits the fastest computational speed compared to other existing state-of-the-art algorithms.
Event(s)
8th International Conference on Imaging, Signal Processing and Communications, ICISPC 2024
SDGs
Publisher
IEEE
Type
conference paper