摘要
为解决K-SVD图像去噪算法运算复杂、去噪效率低的问题,提出一种基于平均计算的快速K-SVD图像去噪算法。采用分簇去噪的思想,按照灰度方差将图像分为背景簇和内容簇两部分;用平均计算预处理策略消除噪声对图像分簇精度的影响;用均值滤波对背景簇去噪,K-SVD算法对内容簇去噪,结合均值滤波的去噪速度快以及K-SVD算法去噪效果好的优势。实验结果表明,该算法保留了K-SVD去噪效果好的优势,在去噪效率上较原算法有明显的改善。
To solve problems of complex matrix operations and inefficiency of K singular value decomposition in image denoising,a mean based speeded-up K-SVD algorithm which improved the efficiency was proposed.Thought of clustering denoising was presented to divide an image into two clusters according to gray variance,the smooth one which was denoised using mean filter and context one which was denoised using K-SVD were taken advantage of.Results of extensive experiments show the proposed method achieves the state-of-the-art denoising performance.
出处
《计算机工程与设计》
北大核心
2017年第12期3380-3384,共5页
Computer Engineering and Design
基金
国家自然科学基金项目(61673249
61273291)
山西省重点实验室开放课题基金项目(2016002)
山西省社科联重点课题基金项目(SSKLZDKT2017126)
忻州师院科研基金项目(201705)