Image Denoising Algorithm Based on Nonlocal Regularization Sparse Representation

Abstract

Image denoising is essential for subsequent image processing and application, and it ensures that people can obtain valid information of images more accurately. The key to image denoising is how to preserve the structure and detail information in the original image while removing noise. By the further analysis of image features, based on K-SVD image denoising algorithm, an image sparse regularization denoising algorithm based on structural similarity is proposed. In this paper, the structural similarity is used instead of the mean square error as the fidelity term in the sparse regularization denoising model. By using it as an optimization criterion, the geometric features of the image can be restored to the utmost. The experimental results show that compared with the K-SVD image denoising algorithm, the algorithm can better preserve the structural information of the image and obtain better image visual effects while effectively denoising.