MRMark：面向三维视频融合的机器性能评测方法

doi:10.16182/j.issn1004731x.joss.19-VR0442

系统仿真学报 ›› 2020, Vol. 32 ›› Issue (7): 1375-1384.doi: 10.16182/j.issn1004731x.joss.19-VR0442

• 仿真支撑平台/系统技术 • 上一篇下一篇

MRMark：面向三维视频融合的机器性能评测方法

孙奕飞^{1, 2}, 蔡亮亮¹, 周颐^{1, 2}, 周忠¹

1. 虚拟现实技术与系统国家重点实验室,北京航空航天大学,北京 100191;
2. 北京大视景科技有限公司,北京 100088

收稿日期:2019-08-30 修回日期:2019-11-05 出版日期:2020-07-25 发布日期:2020-07-15
作者简介:孙奕飞(1996-),男,江苏泰州,硕士生,研究方向为虚实融合;蔡亮亮(1996-),男,河南焦作,博士生,研究方向为图像匹配与融合。

MRMark: A Machine Performance Evaluation Method for Three-dimensional Video Fusion

Sun Yifei^{1, 2}, Cai Liangliang¹, Zhou Yi^{1, 2}, Zhou Zhong¹

1. State Key Laboratory of Virtual Reality Technology and Systems, Beihang University, Beijing 100191, China;
2. Beijing Bigview Technology Co., Ltd, Beijing 100088, China

Received:2019-08-30 Revised:2019-11-05 Online:2020-07-25 Published:2020-07-15

摘要/Abstract

摘要： 三维视频融合系统涉及视频和图形的计算,对机器性能有较高要求,但现有商用评测工具缺乏对应考虑,会出现机器评分与实际运行性能不相符的情况。提出一种面向三维视频融合的机器性能评测方法,并实现了评测工具MRMark。该方法根据视频融合系统的特点设计了多组标准三维测试场景,统计测试过程中的帧率、播放的视频路数等参数,来计算GPU、CPU分数和耦合分数,以及机器的整体性能分数。实验结果表明,相比现有评测工具,该工具的评测结果能更有效地检测机器进行三维视频融合计算的能力。

关键词: 性能评测, 评测工具, 虚拟场景, 虚实融合技术

Abstract: The three-dimensional video fusion system needs the calculation of video and graphics, and requires high machine performance. The existing commercial evaluation tools lack the corresponding considerations, which may cause the inconsistency of a machine's performance from the score. A machine performance evaluation method for three-dimensional video fusion is proposed and an evaluation tool MRMark is implemented. Several sets of standard 3D test scenes are designed according to the characteristics of the video fusion system, the statistics on the frame rate, the number of video channels played and the other parameters during the test are collected to calculate the GPU, CPU and coupling score, as well as the overall performance score of the machine. The results show that, compared to the existing evaluation tools, the tool can more effectively detect the ability of the machine to perform the three-dimensional video fusion calculations.

Key words: performance evaluation, evaluation tool, virtual scene, virtual-reality integration

中图分类号:

TP311.56

孙奕飞, 蔡亮亮, 周颐, 周忠. MRMark：面向三维视频融合的机器性能评测方法[J]. 系统仿真学报, 2020, 32(7): 1375-1384.

Sun Yifei, Cai Liangliang, Zhou Yi, Zhou Zhong. MRMark: A Machine Performance Evaluation Method for Three-dimensional Video Fusion[J]. Journal of System Simulation, 2020, 32(7): 1375-1384.

参考文献 8

[1]	Hsu S, Samarasekera S, Kumar R, et al.Pose estimation, model refinement, and enhanced visualization using video[C]. IEEE Conference on Computer Vision and Pattern Recognition. USA: IEEE, 2000, 1: 488-495.
[2]	Sawhney H S, Arpa A, Kumar R, et al.Video flashlights: real time rendering of multiple videos for immersive model visualization[C]. Proceedings of the 13th Eurographics workshop on Rendering. Goslar, Germany: Eurographics Association, 2002: 157-168.
[3]	Neumann U, You S, Hu J, et al.Augmented virtual environments(AVE): Dynamic fusion of imagery and 3d models[C]. Proceedings of IEEE Virtual Reality. USA: IEEE Computer Society, 2003: 61.
[4]	Hu J H.Intergrating complementary information for photorealistic representation[D]. Los Angeles: University of Southern California, 2009.
[5]	Decamp P, Shaw G, Kubat R, et al.An immersive system for browsing and visualizing surveillance video[C]. Acm International Conference on Multimedia. USA: ACM, 2010.
[6]	Pan C, Chen Y, Wang G.Virtual-Real Fusion with Dynamic Scene from Videos[C]. 2016 International Conference on Cyberworlds (CW). USA: IEEE Computer Society, 2016.
[7]	邓磊, 陈宝华, 赖伟良, 等. 三维监控系统中基于三维重构的交互式标定[J]. 电子学报, 2017, 45(3): 527-533.Deng Lei, Chen Baohua, Lai Weiliang, et al.Interactive Calibrating 3D Surveillance System based on 3D Reconstruction[J]. Acta Electronica Sinica, 2017, 45(3): 527-533.
[8]	周颐, 孟明, 吴威, 等. 基于视频模型的虚拟现实视频融合系统[J]. 系统仿真学报, 2018, 30(7): 2550-2257.Zhou Yi, Meng Ming, Wuwei, et al. Virtual-reality Video Fusion System Based on Video Model[J]. Journal of System Simulation, 2018, 30(7): 2550-2557.

MRMark：面向三维视频融合的机器性能评测方法

MRMark: A Machine Performance Evaluation Method for Three-dimensional Video Fusion

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献 8

相关文章 3

编辑推荐

Metrics

本文评价

[1]	石敏, 魏育坤, 金相臣, 王素琴, 毛天露. 可支持跃层寻径的虚拟场景漫游方法[J]. 系统仿真学报, 2019, 31(7): 1358-1366.
[2]	周颐, 孟明, 吴威, 周忠. 基于视频模型的虚拟现实视频融合系统[J]. 系统仿真学报, 2018, 30(7): 2550-2557.
[3]	孙智仲, 卢泽辉, 李蔚清. 基于虚拟场景距离测度的复合卡尔曼运动估计[J]. 系统仿真学报, 2016, 28(10): 2534-2539.