{"title":"Distributed Video Transcoding System for 8K 360° VR Tiled Streaming Service","authors":"Yong-Hwan Kim, Jingang Huh, Jinwoo Jeong","doi":"10.1109/ICTC.2018.8539372","DOIUrl":null,"url":null,"abstract":"360° Virtual Reality (VR) services with resolutions of 8K and beyond are a challenging task due to limits of both decoding complexity and constrained public internet bandwidth of consumer devices. Also, general streaming servers cannot service these large-resolution video streams to many clients because of bandwidth limitation. In this paper, we propose a distributed video transcoding system for achieving viewport adaptive streaming, which is known as tiled streaming, of 8K 360° VR video. The proposed system consists of many motion-constrained High Efficiency Video Coding (HEVC) encoders, a Hadoop/Spark-based distributed computing platform, light-weight bitstream stitcher, and dual HEVC decoders. Experimental results show that 8K 360° videos which are split by 8×8 tiles, respectively, can be encoded at 99 fps, and 4×4 tiles are stitched at 9,585 fps, on average.","PeriodicalId":417962,"journal":{"name":"2018 International Conference on Information and Communication Technology Convergence (ICTC)","volume":"106 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 International Conference on Information and Communication Technology Convergence (ICTC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICTC.2018.8539372","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
360° Virtual Reality (VR) services with resolutions of 8K and beyond are a challenging task due to limits of both decoding complexity and constrained public internet bandwidth of consumer devices. Also, general streaming servers cannot service these large-resolution video streams to many clients because of bandwidth limitation. In this paper, we propose a distributed video transcoding system for achieving viewport adaptive streaming, which is known as tiled streaming, of 8K 360° VR video. The proposed system consists of many motion-constrained High Efficiency Video Coding (HEVC) encoders, a Hadoop/Spark-based distributed computing platform, light-weight bitstream stitcher, and dual HEVC decoders. Experimental results show that 8K 360° videos which are split by 8×8 tiles, respectively, can be encoded at 99 fps, and 4×4 tiles are stitched at 9,585 fps, on average.