摘要
本文从不同的角度考虑OSV模型,提出一种基于全变差和H-1范数的图像分解变分模型.通过分析OSV模型的性质,给出该模型基于小波空间的非线性偏微分方程和迭代算法.同时,从理论上分析了该模型的极小值存在性.实验表明该方法具有可行性.
A variational model for image decomposition based on total variation and H^-1 norm is proposed, whose start is different from the OSV model.By the propertied of OSV model,the nonlinear partial differential equation and the associated iterafive algorithm based on wavelet method are introduced. And the proof of the existence of minimizer for the variational model is given.Numerical results of image decomposition and denoising show that this model is feasible.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2008年第1期184-187,共4页
Acta Electronica Sinica
关键词
全变差
图像分解
结构
纹理
偏微分方程
极小化
小波
total variation
image decomposition
structure
texture
partial differential equation
minimization
wavelet