一种基于全变分模型的图像修复改进算法

Improved image inpainting algorithm based on total variationmodel

为了恢复图像中划痕、文字等小目标去除后丢失的相关信息,对全变分(TV)模型及其自适应算法进行了分析和改进。在Chan提出的图像修复原则的基础上给出了两个阈值参数,对原有算法中的权值系数进行了改进。仿真实验结果表明,本文算法在保证原有算法修复效果的同时能够有效地提高运算速度,取得了较好的实际效果。

In order to recover the missing information after removing small objects from an image,such as the scratches and characters,we analyze the total variation（TV） model and its self-adaptive algorithm in this paper.Based on the image inpainting principles proposed by Tony F.Chan,the weight coefficient in the original algorithm is improved with two proposed threshold parameters.Simulation results show that the proposed algorithm can improve the computing speed effectively,and at the same time,can keep the similar inpainting effect to the original one.