Group Sparsity Residual with Non-Local Samples for Image Denoising

Abstract

Inspired by group-based sparse coding, recently proposed group sparsity residual (GSR) scheme has demonstrated superior performance in image processing. However, one challenge in GSR is to estimate the residual by using a proper reference of the group-based sparse coding (GSC), which is desired to be as close to the truth as possible. Previous researches utilized the estimations from other algorithms (i.e., GMM or BM3D), which are either not accurate or too slow. In this paper, we propose to use the Non-Local Samples (NLS) as reference in the GSR regime for image denoising, thus termed GSR-NLS. More specifically, we first obtain a good estimation of the group sparse coefficients by the image nonlocal self-similarity, and then solve the GSR model by an effective iterative shrinkage algorithm. Experimental results demonstrate that the proposed GSR-NLS not only outperforms many state-of-the-art methods, but also delivers the competitive advantage of speed.

Image denoising by low‐rank approximation with estimation of noise energy distribution in SVD domain

“…In this section, we validate the superiority of our method over several state‐of‐the‐art methods, including ASCPCA [26], LRA‐SVD [37], Adaptive denoising method [11] and GSR‐NLS [28]. We also use PSNR and SSIM to evaluate the quality of the denoised images.…”

“…The quantitative experimental results are shown in Table 2, where the PSNR and SSIM are computed for all competing methods, our method and benchmark methods. Table 2 shows that our method outperforms ASCPCA [26], LRA‐SVD [37], adaptive denoising method [11] and GSR‐NLS [28] in a wide range of noise levels. For some images including Image Lena and Image Zelda , our method achieves considerable improvement over the other methods.…”

“…Efficiency is another key factor in evaluating an algorithm. We then evaluate the computation cost by comparing the running time of BM3D [20], ASCPCA [26], SAIST [36], WNNM [34], LRA‐SVD [37], adaptive denoising method [11], GSR‐NLS [28] and the proposed method on a grey image with noise standard deviation 20, 30, 50 and 75. All the source codes are performed on the same platform of an Intel Core i5‐4590 CPU 3.30 GHz with 8GB memory.…”

“…Liu [26] proposed an adaptive sparse coding on PCA dictionary, which learns a dictionary using PCA basis [27] and denoises by adaptive sparse coding. Inspired by the success of the sparse representation, group‐based sparse coding has attracted considerable interests in image denoising [28–30].…”

Adaptive iterative global image denoising method based on SVD

“…Zha et al . [32] combines non‐local samples and iterative shrinking group‐based sparse coding to achieve image denoising. Although the dictionary‐based algorithm [31–33] has a better denoising effect, it takes a long time.…”