Skip to Main Content
An adaptive nonlocal-means (ANL-means) algorithm for image denoising is proposed in this work. It employs the singular value decomposition (SVD) method and the K-means clustering (K-means) technique to achieve robust block classification in noisy images. Then, a local window is adaptively adjusted to match the local property of a block and a rotated matching algorithm that aligns the dominant orientation of a local region is adopted for similarity matching. Furthermore, the noise level is estimated using the block classification result and the Laplacian operator. Experimental results are given to demonstrate the superior denoising performance of the proposed ANL-means denoising technique over various image denoising benchmarks in terms of the PSNR value and perceptual quality comparison, where images corrupted by additive white Gaussian noise (AWGN) are tested.