基于格子波尔兹曼的图像修复模型研究  

Study of Lattice Boltzmann Method Based Image Inpainting

作　　者：刘作楚[1] 陈翰林[2] 鲜大权[2] 

机构地区：[1]西南科技大学信息工程学院,四川绵阳621010 [2]西南科技大学信理学院,四川绵阳621010

出　　处：《西南科技大学学报》2013年第2期64-68,共5页Journal of Southwest University of Science and Technology

基　　金：国家自然科学基金项目(10971169)

摘　　要：在研究格子波尔兹曼图像修复模型的基础上,提出了一种改进的模型,该模型通过引入调节函数,使图像在梯度变化较大的区域减弱其扩散强度,在梯度变化较小的区域增强其扩散强度,同时引入调节因子来控制其作用强度。实验结果证明改进后的模型具有更好的修复效果。A Lattice-Boltnnann method （LBM） based image inpainting was introduced in detail and then an image inpainting algorithm was proposed,which was improved by introducing a constraint function.This constraint function can not only weaken the intensity of the diffusion while the image gradient changes quickly,but also enhance the strength of the diffusion while the image gradient changes slowly.What＇s more,a regulatory factor was introduced to control the intensity of the diffusion conveniently.Experimental results show that the proposed algorithm has better general performance in inpainting quality.

关 键 词：整体变分 曲率驱动扩散 非线性各向异性扩散 格子玻尔兹曼法 边缘停止函数 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]