全变差图像恢复的自适应步长梯度投影算法

针对图像去噪问题,本文基于全变差对偶公式提出一个新的梯度投影算法.算法采用改进的非单调线搜索和自适应BB（Barzilai-Borwein）步长,有效地改善了Chambolle梯度投影算法收敛慢的缺点.数值结果表明新算法优于一些已有的梯度投影算法....

Full description

Saved in:
Bibliographic Details
Published in自动化学报 Vol. 42; no. 9; pp. 1347 - 1355
Main Author 张本鑫 朱志斌
Format Journal Article
LanguageChinese
Published 桂林电子科技大学广西高校数据分析与计算重点实验室 桂林 541004 2016
桂林电子科技大学广西自动检测技术与仪器重点实验室 桂林 541004%桂林电子科技大学数学与计算科学学院 桂林 541004
桂林电子科技大学电子工程与自动化学院 桂林 541004
Subjects
全变差
图像恢复
改进的线搜索
梯度投影
自适应步长
改进的线搜索
全变差
image restoration
梯度投影
自适应步长
adaptive steplength selection
图像恢复
total variation
new line search
Gradient pro jection
Online AccessGet full text
ISSN0254-4156
1874-1029
DOI10.16383/j.aas.2016.c150146

Cover

More Information
Summary:针对图像去噪问题,本文基于全变差对偶公式提出一个新的梯度投影算法.算法采用改进的非单调线搜索和自适应BB（Barzilai-Borwein）步长,有效地改善了Chambolle梯度投影算法收敛慢的缺点.数值结果表明新算法优于一些已有的梯度投影算法.
Bibliography:We propose a new gradient projection algorithm for image denoising based on the dual of total variation. The new method exploits nonmonotone line-search and adaptive steplength selection based on strategies for alternation of the well-known Barzilai-Borwein rules. The proposed method is much faster than the Chambolle＇s gradient projection algorithm. Numerical results illustrate the efficiency of this method.
ZHANG Ben-Xin,ZHU Zhi-Bin （1. School of Electronic Engineering and Automation, Guilin University of Electronic Technology, Guilin 541004 ;2. Guangxi Key Laboratory of Automatic Detecting Technology and Instruments, Guilin University of Electronic Technology, Guilin 541004 ;3. School of Mathematics and Computing Science, Guilin University of Electronic Technology, Guilin 541004 ;4. Guangxi Colleges and Universities Key Laboratory of Data Analysis and Computation, Guilin University of Electronic Technology, Guilin 541004_
Gradient projection, total variation, adaptive steplength selection, new line search, im
ISSN:0254-4156
1874-1029
DOI:10.16383/j.aas.2016.c150146