基于样本块的图像修复方法  

Image Inpainting Method Based on Sample Patch

作　　者：张君 常霞[1] 王利娟 ZHANG Jun;CHANG Xia;WANG Li-juan(The Key Laboratory of Intelligent Information and Big Data Processing of Ningxia Province,North Minzu University,Yinchuan Ningxia 750021,China)

机构地区：[1]北方民族大学宁夏智能信息与大数据处理重点实验室

出　　处：《科技视界》2019年第33期21-23,3,共4页Science & Technology Vision

基　　金：国家自然科学基金项目(No.61561001);宁夏自然科学基金项目(No.2018AAC03126);宁夏高等学校一流学科建设(数学学科)(No.NXYLXK2017B09);北方民族大学重大专项项目(No.ZDZX201801)资助;北方民族大学研究生创新项目(YCX19132)

摘　　要：图像修复是图像处理的一个重要研究部分,图像修复方法是试图估计原图像信息,对破损区域进行修补和改善,提高图像视觉质量的技术。基于样本块的图像修复方法以修复速度快,处理效果好,且对破损面积较大的图像修复效果佳等显著优点,成为当前图像修复的热点方向之一。本文详细介绍了基于样本块图像修复方法的优缺点和改进方向,为研究者在图像修复上提供了相关的理论依据,同时也为图像修复方法创新提供了研究思路。文章最后给出了图像修复的发展方向,对深入学习图像修复方法具有重要的指导意义。Image inpainting is one of the important contents in image processing. The image inpainting method is a technique that attempts to estimate the original image information, repair and improve the damaged area, and improve the visual quality of the original picture. The image inpainting method based on sample patchs has obvious advantages such as fast restoration speed, good processing effect and good image restoration effect for images with large damage area,becoming one of the hot spots of image restoration at present. In this paper, the advantages, disadvantages and improvement directions of image restoration methods based on sample blocks are introduced in detail, which provides relevant theoretical basis for researchers in image inpainting and research ideas for image restoration method innovation.At the end of this paper, the development direction of image restoration is given, which has important guiding significance for further study of image restoration methods.

关 键 词：图像修复 样本块 优先权 置信度 数据项 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]