基于K-SVD算法改进BayesShrink小波阈值去噪

Using K-SVD algorithm for improving performance of BayesShrink image denoisng techniques

提出了一种基于BayesShrink小波阈值去噪算法和稀疏字典学习算法(K-SVD)相结合的图像去噪算法。针对现有的小波去噪算法只处理了细节子带系数,而没有处理近似子带的系数最终导致去噪效果带有局限性的问题,在实际应用中,噪声不仅改变了细节子带系数同时还改变了近似子带的系数,提出了使用K-SVD算法处理图像小波变换近似子带系数以改进现有小波阈值图像去噪算法的效果的缺陷,仿真实验结果表明:改进后的算法能够有效的去除图像的高斯噪声,提高图像的峰值信噪比,明显的改善图像的视觉效果。

A new method based on K- SVD and wavelel transfoml is presented tbr image noise reduction. In the existing wavelet thresholding methods, the final noise reduced image has limited improvement. It is due to keeping the approximate coefficients of the image unchanged. These coefficients have the main information of the image. Since noise affects both the approximate and detail coefficients, in this research, the K- SVD technique for noise reduction is applied on the approximation band to alleviate the deficiency of the existing wavelet thresholding methcds. The proposed method was applied on several standard noisy inmges and the results indicate superiority of the proposed method over the existing wavelet - based image denoising, BayesShrink, Fuzzy - Shrink, and AntShrink method, the proposed method is obviously superior both in vision and in PSNR.