{"title":"Binary tree decomposition depth coding for 3D video applications","authors":"Gonçalo Carmo, M. Naccari, F. Pereira","doi":"10.1109/ICME.2011.6012254","DOIUrl":null,"url":null,"abstract":"Recent advances in three dimensional display technologies and the growing efforts put in the production of three dimensional videos are intensely demanding for new, more efficient three dimensional content representation formats. One promising format is the so-called multiview plus depth format where multiple views of the observed scene are represented together with a per pixel depth map. Naturally, this format requires to efficiently code not only the video views but also the depth data. In this context, this paper proposes a novel depth map codec encoding the depth data by means of a binary tree triangular decomposition and reconstructing the depth map values by means of a triangle based planar approximation. For depth maps related to typical three dimensional video contents, the proposed depth map codec outperforms both the H.264/AVC standard with all intra coding modes enabled and the JPEG standard. In particular, for the same objective reconstruction quality, the proposed codec allows an average bitrate reduction of 60% and 90% regarding H.264/AVC Intra and JPEG coding, respectively.","PeriodicalId":433997,"journal":{"name":"2011 IEEE International Conference on Multimedia and Expo","volume":"48 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2011-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2011 IEEE International Conference on Multimedia and Expo","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICME.2011.6012254","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 6
Abstract
Recent advances in three dimensional display technologies and the growing efforts put in the production of three dimensional videos are intensely demanding for new, more efficient three dimensional content representation formats. One promising format is the so-called multiview plus depth format where multiple views of the observed scene are represented together with a per pixel depth map. Naturally, this format requires to efficiently code not only the video views but also the depth data. In this context, this paper proposes a novel depth map codec encoding the depth data by means of a binary tree triangular decomposition and reconstructing the depth map values by means of a triangle based planar approximation. For depth maps related to typical three dimensional video contents, the proposed depth map codec outperforms both the H.264/AVC standard with all intra coding modes enabled and the JPEG standard. In particular, for the same objective reconstruction quality, the proposed codec allows an average bitrate reduction of 60% and 90% regarding H.264/AVC Intra and JPEG coding, respectively.