整体变分模型在矢量图像修复中的应用  

Color image inpainting based on total variation model

作　　者：张福美[1] 

机构地区：[1]莱芜职业技术学院计算机系,山东莱芜271100

出　　处：《计算机应用》2008年第4期993-994,998,共3页journal of Computer Applications

摘　　要：图像修复是指恢复图像中破损区域的颜色信息或者去除图像中的多余物体。分析了基于整体变分法TV模型以及矢量图像耦合技术的原理,根据矢量图像耦合思想将整体变分法运用到矢量图像中并对矢量图像进行试验。实验结果表明:改进的矢量图像耦合修复模型能较好地修复大块彩色图像的缺失信息和移除多余物体,能保持彩色图像的边缘,且有较好的去噪功能。The goal of digital image inpainting is to restore damaged regions or remove objects in the image. Both Tural Variation （TV） method and a basic coupled method for color image were analyzed. According to a basic coupled method for color image, color images were processed based on the total variation method, which was used in inpainting gray images, Experimental results show that the improved channel coupled diffusion model is better in inpainting lost block information and removing odd objects, keeps better edges, and is effective for image denoising in color image.

关 键 词：图像修复 彩色图像耦合 整体变分法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]