A fast nonlocally centralized sparse representation algorithm for image denoising
详细信息 查看全文
- 作者：Shaoping Xu ; xushaoping@ncu.edu.cn" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Xiaohui Yang yangxiaohui@ncu.edu.cn" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Shunliang Jiang jiangshunliang@ncu.edu.cn" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae
- 关键词：Image denoising ; Nonlocally centralized sparse representation ; Computational efficiency ; Pre-learned dictionary ; Adaptive parameter setting
- 刊名：Signal Processing
- 出版年：2017
- 出版时间：February 2017
- 年：2017
- 卷：131
- 期：Complete
- 页码：99-112
- 全文大小：3225 K
文摘
- •
A fast version of the nonlocally centralized sparse representation (FNCSR) algorithm is proposed.
- •
We utilize the pre-learned dictionary instead of the adaptive dictionary constructed runtime.
- •
We utilize the training-based NLE for automatic and optimal parameter setting.
- •
The thorough quantitative and qualitative results demonstrate that FNCSR achieves better results than state-of-the-art algorithms.