基于剪切波的变分图像放大方法

›› 2016, Vol. 29 ›› Issue (6): 146-.

基于剪切波的变分图像放大方法

王鹏，吴玉莲

（1中航工业西安航空计算技术研究所 13室，陕西西安 710065；2西安医学院卫生管理系，陕西西安 710021）

出版日期:2016-06-15 发布日期:2016-06-22
作者简介:王鹏（1978-），男，工程师。研究方向：图像处理。
基金资助:
陕西省教育厅基金资助项目（15JK1371）；西安医学院博士科研启动基金资助项目（2015DOC25）

Combining Variation and Shearlet Transform for Image Zooming

WANG Peng，WU Yulian

(1No13 Room, AVIC Xi'an Aeronautical Engineering Institute of Computing Technology, Xi'an 710065, China; 
2 Department of Health ,Xi’an Medical College, Xi’an 710021，China)

Online:2016-06-15 Published:2016-06-22

摘要/Abstract

摘要：

针对图像放大的Chambolle 变分模型会出现阶梯效应的现象，文中提出了一种基于Shearlet光滑分解空间的变分模型。利用有界变差空间和Shearlet分解空间的关系，特别是Shearlet分解空间的半范与加权Shearlet系数之间的等价关系,将所求的变分问题转化为基于Shearlet域的变分问题,其解归结于简单的Shearlet阈值。实验仿真表明，该方法放大后的图像有效地消除了阶梯块效应，保持了更多的细节，具有更高的峰值信噪比。

关键词: 图像放大, 剪切波, 变分模型, 分解空间

Abstract:

As the Chambolle variational model of image magnification results in the staircase effect. a new algorithm for image zooming based on shearlet smoothness decomposition spaces is proposed. The new algorithm translates the variational problem that is solved into a sequence based shearlet field through the equivalent relationship between semi-norm of shearlet-type decomposition spaces and the weighted shearlet coefficients, solution of the proposed model approximately equals to different shearlet shrinkages. Experimental results have verified that the new algorithm can effectively eliminate the staircase effect of zoomed images and keep more details with a higher peak signal to noise ratio.

Key words: image zooming, Shearlet, variational model, decomposition spaces

中图分类号:

TP391.41

王鹏，吴玉莲. 基于剪切波的变分图像放大方法[J]. , 2016, 29(6): 146-.

WANG Peng，WU Yulian. Combining Variation and Shearlet Transform for Image Zooming[J]. , 2016, 29(6): 146-.