[1]张晓严,马杰,武利涛. 基于过完备字典学习的全变分图像去噪方法[J].河北工业大学学报,2017,(01):1-7.[doi:10.14081/j.cnki.hgdxb.2017.01.001]
 ZHANG Xiaoyan,MA Jie,WU Litao. A total variation image denoising method over complete dictionary learning[J].Journal of Hebei University of Technology,2017,(01):1-7.[doi:10.14081/j.cnki.hgdxb.2017.01.001]
点击复制

 基于过完备字典学习的全变分图像去噪方法()
分享到:

《河北工业大学学报》[ISSN:1007-2373/CN:13-1208/T]

卷:
期数:
2017年01期
页码:
1-7
栏目:
出版日期:
2017-02-25

文章信息/Info

Title:
  A total variation image denoising method over complete dictionary learning
文章编号:
1007-2373(2017) 01-0001-08
作者:
 张晓严1 马杰12 武利涛1
1.河北工业大学电子信息工程学院, 天津300401; 2.天津市电子材料与器件重点实验室, 天津300401
Author(s):
 ZHANG Xiaoyan1MA Jie12 WU Litao1
1.School of Electronics and Information Engineering,Hebei University of Technology,Tianjin 300401,China;2.OKey Laboratory of Tianjin Electronic Materials and Devices,Tianjin 300401,China
关键词:
 字典学习 图像去噪 增广拉格朗日乘子法 正交匹配追踪 全变分
Keywords:
 dictionary learning image denoising augmented lagrange multiplier orthogonal matching pursuit total variation
分类号:
TP391.9
DOI:
10.14081/j.cnki.hgdxb.2017.01.001
文献标志码:
A
摘要:
 为了改善低信噪比情况下去噪效果、边缘保持能力差的问题, 提出一种联合全变分正则项的字典学习图像去噪方法. 首先, 把增广拉格朗日乘子法和正交匹配追踪这两种求解稀疏编码的方法跟经典的K-SVD 思想相交融, 改善字典性能; 其次, 将全变分去噪模型融入到基于字典学习的图像去噪理论中, 在图像重构基础上,引入全变分约束项, 作为改进去噪模型中新的一项, 达到对噪声和图像边缘作后续优化处理、改善图像去噪性能的目的. 实验结果表明, 改进的去噪方法, 在保持原有去噪效果前提下, 在噪声标准差较大或者图片边缘信息丰富时, 去噪图像更加自然, 边缘更加清晰, 视觉效果较好.
Abstract:

Abstract In order to improve the low SNR on denoising ability of edge preserving image, proposed K-SVD dictionary learning and total variation regularization denoising method. Firstly, the augmented Lagrange multiplier method and or-thogonal method of the two kinds of solving sparse encoding tracking matching with K-SVD classic thoughts to improve theperformance of the dictionary; the total variational denoising model into image dictionary learning denoising theory improved image denoising performance. The experimental results show that the improved denoising method, while maintaining the original denoising effect, the noise standard deviation is larger or for the rich image edge information, image denoising is more natural with clearer edge, and the visual effect is better.

相似文献/References:

[1]王 停,唐海林,赵宗佑,等. 一种高增益大带宽的增益自举型折叠共源共栅放大器设计[J].河北工业大学学报,2016,(04):20.[doi:10.14081/j.cnki.hgdxb.2016.04.004]
 WANG Ting,TANG Hailin,ZHAO Zongyou,et al. Design on a gain-boosted cascode amplifier with high gain and highunity-bandwidth[J].Journal of Hebei University of Technology,2016,(01):20.[doi:10.14081/j.cnki.hgdxb.2016.04.004]
[2]张婷婷,于 明,阎 刚. 运动目标追踪的人工鱼群优化粒子滤波算法[J].河北工业大学学报,2016,(04):12.[doi:10.14081/j.cnki.hgdxb.2016.04.003]
 ZHANG Tingting,YU Ming,YAN Gang. Moving object tracking based on artificial fish particle filtering[J].Journal of Hebei University of Technology,2016,(01):12.[doi:10.14081/j.cnki.hgdxb.2016.04.003]
[3]向 敏,张艳梅,冯 冲. 预应力管道冻胀引起梁体应力场有限元分析[J].河北工业大学学报,2016,(04):86.[doi:10.14081/j.cnki.hgdxb.2016.04.014]
 XIANG Min,ZHANG Yanmei,FENG Chong. Finite element analysis of beam stress field caused by prestressed pipe frost heaving[J].Journal of Hebei University of Technology,2016,(01):86.[doi:10.14081/j.cnki.hgdxb.2016.04.014]
[4]张晓帆,伍萍辉,杨泽明,等. 基于特征统计的木马检测系统的设计与实现[J].河北工业大学学报,2016,(06):9.[doi:10.14081/j.cnki.hgdxb.2016.06.002]
 ZHANG Xiaofan,WU Pinghui,YANG Zeming,et al. Design and implementation of trojan detection system based on characteristics statistics[J].Journal of Hebei University of Technology,2016,(01):9.[doi:10.14081/j.cnki.hgdxb.2016.06.002]
[5]宋云燕,郭志涛,田学民,等. 一种安卓平台下光伏发电站数据监控系统[J].河北工业大学学报,2016,(06):16.[doi:10.14081/j.cnki.hgdxb.2016.06.003]
 SONG Yunyan,GUO Zhitao,TIAN Xuemin,et al.Study of photovoltaic power station data monitoring and control system based on the android[J].Journal of Hebei University of Technology,2016,(01):16.[doi:10.14081/j.cnki.hgdxb.2016.06.003]
[6]梁晨光,高振斌,白 明,等. 无线嵌入式Web服务器portal认证设计实现[J].河北工业大学学报,2016,(06):22.[doi:10.14081/j.cnki.hgdxb.2016.06.004]
 LIANG Chenguang,GAO Zhenbin,BAI Ming,et al. Design and implementation of portal authentication on wireless embedded web server[J].Journal of Hebei University of Technology,2016,(01):22.[doi:10.14081/j.cnki.hgdxb.2016.06.004]
[7]马 杰,孙耀杰. 基于混合定位的学生校园活动评价系统设计[J].河北工业大学学报,2016,(06):28.[doi:10.14081/j.cnki.hgdxb.2016.06.005]
 MA Jie,SUN Yaojie. Design the evaluation system of students in campus activities based on the mixed localization[J].Journal of Hebei University of Technology,2016,(01):28.[doi:10.14081/j.cnki.hgdxb.2016.06.005]
[8]王博文,崔晓静,李亚芳,等. 磁致伸缩驱动器的结构设计与输出特性研究[J].河北工业大学学报,2016,(05):1.[doi:10.14081/j.cnki.hgdxb.2016.05.001]
 WANG Bowen,CUI Xiaojing,LI Yafang,et al. Structure design and output characteristic analysis of magnetostrictive actuator[J].Journal of Hebei University of Technology,2016,(01):1.[doi:10.14081/j.cnki.hgdxb.2016.05.001]
[9]顾军华,赵 燕,董 瑶,等. 基于FWA-SVM的室内无线定位研究[J].河北工业大学学报,2016,(06):35.[doi:10.14081/j.cnki.hgdxb.2016.06.006]
 GU Junhua,ZHAO Yan,DONG Yao,et al. Wireless indoor location method based on FWA-SVM[J].Journal of Hebei University of Technology,2016,(01):35.[doi:10.14081/j.cnki.hgdxb.2016.06.006]
[10]张维强,李永建,薛 刚,等. 1 kW推挽升压开关电源设计[J].河北工业大学学报,2016,(05):7.[doi:10.14081/j.cnki.hgdxb.2016.05.002]
 ZHANG Weiqiang,LI Yongjian,XUE Gang,et al. The design of the 1 kW push-pull switching power supply[J].Journal of Hebei University of Technology,2016,(01):7.[doi:10.14081/j.cnki.hgdxb.2016.05.002]
[11]刘雅莉,马 杰,王晓云,等. 一种改进的K-SVD字典学习算法[J].河北工业大学学报,2016,(02):1.[doi:10.14081/j.cnki.hgdxb.2016.02.001]
 LIU Yali,MA Jie,WANG Xiaoyun,et al. An improved K-SVD dictionary learning algorithm[J].Journal of Hebei University of Technology,2016,(01):1.[doi:10.14081/j.cnki.hgdxb.2016.02.001]

更新日期/Last Update: 2017-04-14