|||
Image Inpainting Algorithm of CDD Based on Split Bregman Method
摘要 本文提出了一种基于分裂Bregman(Split Bregman)方法的快速曲率驱动图像修补模型(CDD)算法。由于CDD模型中曲率项的影响,数值求解高阶偏微分方程过程中需要大量迭代运算,修复速度缓慢。鉴于Split Bregman方法在l1正则化问题的成功应用,为提高算法计算速度,本文在CDD修补模型中引入Split Bregman方法。实验结果表明,与其他类似方法相比,本文算法的实现速度显著加快,且视觉效果好。
关键词 图像处理;图像修补;Split Bregman; TV模型;CDD模型
Abstract A fast inpainting algorithm of Curvature Driven Diffusions (CDD)
model is proposed. It is will known that the curvature term makes a large
number of iterations to solve the high order PDE in original CDD model. Due
to the success of the Split Bregman method in l1 regularization problem, the
Split Bregman method is introduced in this paper into the CDD model to speed
up the algorithm. The experiments show that our new algorithm is much faster
and with better vision effects.
Keywords Image processing; Image inpainting; Split Bregman; TV model; CDD model
Archiver|手机版|科学网 ( 京ICP备07017567号-12 )
GMT+8, 2024-4-28 09:01
Powered by ScienceNet.cn
Copyright © 2007- 中国科学报社