Image Denoising Using Green Channel Prior
Zhaoming Kong, Fangxi Deng, Xiaowei Yang
TL;DR
The paper tackles real-world image denoising where noise levels vary with local content and across channels, noting that the green channel often yields higher SNR due to Bayer sampling. It introduces GCP-ID, a lightweight patch-based denoiser that uses a green-channel guided patch search, RGGB patch representation with nonlocal t-SVD, and a CNN-based noise estimator to adapt to content and noise level across raw, sRGB, video, and hyperspectral data. Key contributions include integrating the green channel prior into a unified denoising framework, an efficient RGGB-tensor approach for patch groups, and a CNN-based sigma estimator that reduces manual tuning and enhances adaptivity, with ablation and customization strategies demonstrated. The method achieves competitive performance on real-world datasets while offering computational efficiency and practical extension to HSI, making it suitable for diverse imaging pipelines without requiring ground-truth noise-free data.
Abstract
Image denoising is an appealing and challenging task, in that noise statistics of real-world observations may vary with local image contents and different image channels. Specifically, the green channel usually has twice the sampling rate in raw data. To handle noise variances and leverage such channel-wise prior information, we propose a simple and effective green channel prior-based image denoising (GCP-ID) method, which integrates GCP into the classic patch-based denoising framework. Briefly, we exploit the green channel to guide the search for similar patches, which aims to improve the patch grouping quality and encourage sparsity in the transform domain. The grouped image patches are then reformulated into RGGB arrays to explicitly characterize the density of green samples. Furthermore, to enhance the adaptivity of GCP-ID to various image contents, we cast the noise estimation problem into a classification task and train an effective estimator based on convolutional neural networks (CNNs). Experiments on real-world datasets demonstrate the competitive performance of the proposed GCP-ID method for image and video denoising applications in both raw and sRGB spaces. Our code is available at https://github.com/ZhaomingKong/GCP-ID.