结合分块噪声估计的字典学习图像去噪算法

图像去噪是图像处理领域的重要环节,也是对图像进行后续处理的基础。近年来K-SVD字典学习去噪算法因其耗时短、去噪效果好的特点得到广泛关注和应用。但该算法的适用条件为图像的噪声为加性噪声且噪声标准差已知。针对这一情况,本文先提出一种平滑图像块筛选方法,并将其与奇异值分解(Singular Value Decomposition, SVD)相结合实现对图像的噪声标准差估计。再将得到的噪声估计方法与K-SVD字典学习去噪算法结合起来,提出一种具备噪声估计特性的K-SVD字典学习去噪算法。对多种图像的去噪实验结果表明,与Donoho小波软阈值去噪算法、全变分(Total Variation, TV)去噪算法相比,本文算法不仅能够使去噪后图像的峰值信噪比提升1~3dB,并且能较好地保留图像的细节信息和边缘特征。

Wang Haoran, Xia Kewen, Niu Wenjia, Ren Miaomiao, Li Chuo. An Image Denoising Algorithm Combined with Dictionary Learning and blocked-based Noise Estimation. Computer Engineering and Application

Image denoising is a critical part in the field of image processing, and it is also the basis for the subsequent processing of the image. In recent years, the K-SVD dictionary learning denoising algorithm has been widely concerned and applied because of its short time consuming and outstanding performance. But the application of this algorithm requires the noise in image is additive noise and standard deviation of the noise is known. In view of this situation, this paper proposes a method to select the smooth image blocks and combines it with the Singular Value Decomposition (SVD) to achieve the estimation of the noise standard deviation of the image.Then a new denoising algorithm which has the characteristic of noise estimation is proposed combining with the obtained noise estimation method and the K-SVD dictionary learning denoising algorithm. Experimental results of denoising some images show that, compared with Donoho wavelet soft threshold denoising algorithm and the total variation (TV) denoising algorithm, not only the peak signal to noise ratio(PSNR) of the image denoised by the proposed algorithm is improved by about 1~3dB, but also the detailed information and edge features of the image can be better preserved.