结合网函数插值与TV模型的图像修复算法  

Image Inpainting Based on Net Function Interpolation and TV Model

作　　者：侯海娜[1] 戴芳[1] 杨陈东[2] 

机构地区：[1]西安理工大学理学院,西安710054 [2]西安航空学院理学院,西安710077

出　　处：《计算机系统应用》2016年第12期117-125,共9页Computer Systems & Applications

基　　金：陕西省工业科技攻关项目(2015GY004)

摘　　要：TV(Total Variation)模型用于图像修复时没有考虑缺损区域的方向信息,并且存在收敛速度缓慢以及修复质量较低等问题.针对图像上方向特征明显的条状缺损区域,提出带方向的TV图像修复算法(ADTV).该算法分别针对4种方向(0度、45度、90度、135度)对TV算法离散格式进行改进,并引入方向判断,将缺损区域归类到此4种方向进行修复.实验结果表明,该算法充分利用了条状缺损区域的方向信息,有效提高了图像修复质量.为提高修复效率,将网函数插值分别与TV算法、ADTV算法相结合提出Net-TV算法、Net-ADTV算法.实验结果表明,结合算法不但有效减少了迭代次数,降低了时间成本,加快了收敛速度,而且提高了图像修复效果.TV（Total Variation） model is used for image inpainting without taking the direction of damaged areas into account, while its slow convergent rate and low inpainting quality are either not as good as expected. To handle these problems, Additional Direction Total Variation（ADTV） algorithm is proposed, which aims at the strip damaged areas with obvious direction characteristic in the image. In this algorithm, discrete formats of TV model are improved in corresponding four directions（0°, 45°, 90°, 135°）. The damaged areas will be inpainted in the four types by judging their directions. The experimental results show that the method makes full use of the direction information of the strip damaged areas, and effectively improves the inpainting quality. Finally, in order to improve the efficiency of inpainting, Net-TV algorithm and Net-ADTV algorithm are proposed by combining net function interpolation with TV model and ADTV algorithm respectively. The experimental results show that the combined algorithm not only effectively reduces the number of iterations and time cost, speeds up the convergence rate, but also improves the quality of image inpainting.

关 键 词：TV模型 网函数插值 图像修复 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]