Adaptive non-convex total variation regularisation for image restoration

被引：14

作者：

Fu, S. ^{[1
,2
]}

Zhang, C. ^{[1
]}

机构：

[1] Shandong Univ, Sch Comp Sci & Technol, Jinan 250100, Peoples R China

[2] Shandong Univ, Sch Math, Jinan 250100, Peoples R China

来源：

ELECTRONICS LETTERS | 2010年 / 46卷 / 13期

基金：

中国国家自然科学基金; 中国博士后科学基金;

关键词：

D O I：

10.1049/el.2010.0027

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

An adaptive non-convex total variation regularisation is proposed for blind image restoration. A flux corrected transport (FCT) technique is used to obtain a stable numerical scheme, where a spatially varying constraint allows a better restoration of image edges and fine detail. Finally, its advantages are shown in deblurring edges, denoising and restoring fine details of image simultaneously in experiments.

引用

页码：907 / U56

页数：2

共 6 条

[1]

[Anonymous], 2002, Applied Mathematical Sciences

[2]

[Anonymous], IMAGE PROCESSING ANA

[3]

[Anonymous], 1977, SOLUTIONS ILL POSED

[4] A feature-dependent fuzzy bidirectional flow for adaptive image sharpening [J].

Fu, Shujun ;

Ruan, Qiuqi ;

Wang, Wenqia ;

Gao, Fuzheng ;

Cheng, Heng-Da .

NEUROCOMPUTING, 2007, 70 (4-6) :883-895

[5] Nonlinear Scale Space with Spatially Varying Stopping Time [J].

Gilboa, Guy .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2008, 30 (12) :2175-2187

[6] NONLINEAR TOTAL VARIATION BASED NOISE REMOVAL ALGORITHMS [J].

RUDIN, LI ;

OSHER, S ;

FATEMI, E .

PHYSICA D, 1992, 60 (1-4) :259-268

← 1 →