基于改进Criminisi算法的破损纺织品文物图像修复

IF 0.9 4区物理与天体物理 Q4 ENGINEERING, ELECTRICAL & ELECTRONIC

激光与光电子学进展 Pub Date : 2023-01-01 DOI:10.3788/lop222378

李奇 Li Qi, 李龙 Li Long, 王卫 Wang Wei, 南蓬勃 Nan Pengbo

{"title":"基于改进Criminisi算法的破损纺织品文物图像修复","authors":"李奇 Li Qi, 李龙 Li Long, 王卫 Wang Wei, 南蓬勃 Nan Pengbo","doi":"10.3788/lop222378","DOIUrl":null,"url":null,"abstract":"为修复破损纺织品文物图像，在Criminisi算法基础上，提出一种改进的基于K-means颜色分割的纺织品文物图像修复算法。根据纺织品文物图像的特点，将RGB图像转化为Lab颜色模型，采用K-means分类器对a*b*层数据基于颜色进行分割处理，对纹样图案边缘进行标定并缩小匹配块搜索区域；引入L值的标准差来表示颜色离散度，对优先权函数以及自适应匹配块进行改进。用所提算法与文献报道的3种算法对自然破损纺织品文物图像和人为破损纺织品图像进行修复，并对修复结果进行评价。实验结果表明，所提算法修复的图像纹理自然、结构合理，峰值信噪比、结构相似性、特征相似性、均方误差值更好。","PeriodicalId":51502,"journal":{"name":"激光与光电子学进展","volume":"6 1","pages":"0"},"PeriodicalIF":0.9000,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"基于改进Criminisi算法的破损纺织品文物图像修复\",\"authors\":\"李奇 Li Qi, 李龙 Li Long, 王卫 Wang Wei, 南蓬勃 Nan Pengbo\",\"doi\":\"10.3788/lop222378\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"为修复破损纺织品文物图像，在Criminisi算法基础上，提出一种改进的基于K-means颜色分割的纺织品文物图像修复算法。根据纺织品文物图像的特点，将RGB图像转化为Lab颜色模型，采用K-means分类器对a*b*层数据基于颜色进行分割处理，对纹样图案边缘进行标定并缩小匹配块搜索区域；引入L值的标准差来表示颜色离散度，对优先权函数以及自适应匹配块进行改进。用所提算法与文献报道的3种算法对自然破损纺织品文物图像和人为破损纺织品图像进行修复，并对修复结果进行评价。实验结果表明，所提算法修复的图像纹理自然、结构合理，峰值信噪比、结构相似性、特征相似性、均方误差值更好。\",\"PeriodicalId\":51502,\"journal\":{\"name\":\"激光与光电子学进展\",\"volume\":\"6 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.9000,\"publicationDate\":\"2023-01-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"激光与光电子学进展\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.3788/lop222378\",\"RegionNum\":4,\"RegionCategory\":\"物理与天体物理\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"ENGINEERING, ELECTRICAL & ELECTRONIC\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"激光与光电子学进展","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.3788/lop222378","RegionNum":4,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}

引用次数: 0

摘要

为修复破损纺织品文物图像，在Criminisi算法基础上，提出一种改进的基于K-means颜色分割的纺织品文物图像修复算法。根据纺织品文物图像的特点，将RGB图像转化为Lab颜色模型，采用K-means分类器对a*b*层数据基于颜色进行分割处理，对纹样图案边缘进行标定并缩小匹配块搜索区域；引入L值的标准差来表示颜色离散度，对优先权函数以及自适应匹配块进行改进。用所提算法与文献报道的3种算法对自然破损纺织品文物图像和人为破损纺织品图像进行修复，并对修复结果进行评价。实验结果表明，所提算法修复的图像纹理自然、结构合理，峰值信噪比、结构相似性、特征相似性、均方误差值更好。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

基于改进Criminisi算法的破损纺织品文物图像修复

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

激光与光电子学进展 Multiple-

CiteScore

1.60

自引率

50.00%

发文量

18006

期刊介绍： Laser & Optoelectronics Progress, the first laser and optoelectronics journal published in China. The main columns include general, lasers and laser optics, fiber optics and optical communications, optical design and fabrication, materials, image processing, imaging systems, optical devices, remote sensing and sensors, atmospheric optics and oceanic optics, diffraction and gratings, atomic and molecular physics, detectors, thin films, ultrafast optics, etc. The journal is included in ESCI, INSPEC, Scopus, CSCD, Chinese Core Journals, Chinese Science and Technology Core Journals, and T2 level of the Classified Catalogue of High Quality Science and Technology Journals in Optical Engineering and Optical Fields, and other databases.