Real-time free-viewpoint DIBR on GPUs for 3DTV systems

2011 IEEE International Conference on Consumer Electronics -Berlin (ICCE-Berlin) Pub Date : 2011-09-29 DOI:10.1109/ICCE-BERLIN.2011.6031881

Germán Bravo, L. Do, S. Zinger, P. D. De with

{"title":"Real-time free-viewpoint DIBR on GPUs for 3DTV systems","authors":"Germán Bravo, L. Do, S. Zinger, P. D. De with","doi":"10.1109/ICCE-BERLIN.2011.6031881","DOIUrl":null,"url":null,"abstract":"Multi-view 3D may succeed stereo 3DTV in multimedia and TV applications. To this end, the MPEG committee has installed a special task force to establish a standard for multi-view 3D coding. One focal point of our research work is on an efficient implementation of the rendering part of such a multi-view 3D system, because it is a computationally expensive task and it determines the final reconstructing quality. Our free-viewpoint DIBR algorithm is implemented with an off-the-shelf GPU that can be integrated in advanced 3DTV systems. We present the principal steps of a representative free-viewpoint DIBR and show the key differences between the reference software and our GPU implementation. One of those differences is the joint execution of signal processing blocks to share memory usage. Using a combination of the highly parallel programming architecture CUDA and a graphics API, we have achieved a real-time performance operating on 1080p HD multi-view video with a rendering quality that is comparable to the software implementation.","PeriodicalId":236486,"journal":{"name":"2011 IEEE International Conference on Consumer Electronics -Berlin (ICCE-Berlin)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2011-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"9","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2011 IEEE International Conference on Consumer Electronics -Berlin (ICCE-Berlin)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCE-BERLIN.2011.6031881","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 9

Abstract

Multi-view 3D may succeed stereo 3DTV in multimedia and TV applications. To this end, the MPEG committee has installed a special task force to establish a standard for multi-view 3D coding. One focal point of our research work is on an efficient implementation of the rendering part of such a multi-view 3D system, because it is a computationally expensive task and it determines the final reconstructing quality. Our free-viewpoint DIBR algorithm is implemented with an off-the-shelf GPU that can be integrated in advanced 3DTV systems. We present the principal steps of a representative free-viewpoint DIBR and show the key differences between the reference software and our GPU implementation. One of those differences is the joint execution of signal processing blocks to share memory usage. Using a combination of the highly parallel programming architecture CUDA and a graphics API, we have achieved a real-time performance operating on 1080p HD multi-view video with a rendering quality that is comparable to the software implementation.

查看原文本刊更多论文

3DTV系统gpu实时自由视点DIBR

多视点3D可以在多媒体和电视应用中取代立体3DTV。为此，MPEG委员会成立了一个特别工作组，以建立多视图3D编码标准。我们研究工作的一个重点是如何有效地实现这种多视图三维系统的渲染部分，因为它是一个计算昂贵的任务，它决定了最终的重建质量。我们的自由视点DIBR算法是由一个现成的GPU实现的，可以集成在先进的3DTV系统中。我们给出了一个典型的自由视点DIBR的主要步骤，并展示了参考软件与我们的GPU实现之间的主要区别。其中一个区别是信号处理块的联合执行以共享内存使用。使用高度并行编程架构CUDA和图形API的组合，我们已经实现了在1080p高清多视图视频上运行的实时性能，其渲染质量与软件实现相当。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

2011 IEEE International Conference on Consumer Electronics -Berlin (ICCE-Berlin)

自引率

0.00%

发文量