{"title":"Improving DIBR technique to resolve foreground color/depth edge misalignment","authors":"W. Lie, Chun-Cheng Yeh, Guo-Shiang Lin","doi":"10.1109/ICMEW.2015.7169759","DOIUrl":null,"url":null,"abstract":"In this paper, we propose a method to overcome flaws in the DIBR-synthesized view due to foreground color/depth edge misalignment. Our method is an improvement of the prior work [1], which is based on background sprite model construction for hole-filling. First, the possible flaw areas in the original view are detected and then an alpha matting technique is used to combine two information sources from the foreground and the background to improve the quality of the synthesized view. Experimental results show that our proposed method is capable of resolving such a kind of flaws regardless of complex or simple background, thus making our view synthesis result close to those synthesized with ideal or flawless foreground depths.","PeriodicalId":388471,"journal":{"name":"2015 IEEE International Conference on Multimedia & Expo Workshops (ICMEW)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2015-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2015 IEEE International Conference on Multimedia & Expo Workshops (ICMEW)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICMEW.2015.7169759","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
In this paper, we propose a method to overcome flaws in the DIBR-synthesized view due to foreground color/depth edge misalignment. Our method is an improvement of the prior work [1], which is based on background sprite model construction for hole-filling. First, the possible flaw areas in the original view are detected and then an alpha matting technique is used to combine two information sources from the foreground and the background to improve the quality of the synthesized view. Experimental results show that our proposed method is capable of resolving such a kind of flaws regardless of complex or simple background, thus making our view synthesis result close to those synthesized with ideal or flawless foreground depths.