Human Motion Measurement System Based on Monocular Videos
ZHANG Ye1,ZHOU Xiao-jing1,YANG Xiao-tong2
1. School of Instrument Science and Engineering, Southeast University, Nanjing, Jiangsu 210096, China
2. Huawei Technologies Co. Ltd, Shenzhen, Guangdong 518129, China
Abstract：In order to solve the problem of low efficiency caused by camera calibration in video-based human motion measurement process, a design scheme of human motion measurement system is proposed for uncalibrated monocular videos. The vanishing points and the vanishing lines of the scene are computed by Maximum Likelihood Estimation,and real-time stature is measured according to the cross ratio theorem. On the basis of height measurement, a method for extracting the feature points of the left and right feet is proposed to measure the pedestrian′s stride. The trajectory, real-time speed and stride frequency of the pedestrian are estimated by tracking the motion of centroid, and the speed measurements are fitted by sinusoidal function to prove the conformity with the objective law of walking. The experimental results show that the relative errors of height and average velocity measurements are no more than 0.84% and 3.72% respectively, which prove that the system is accurate and robust.
［1］魏玮, 王丹丹, 刘命, 等. 人体运动分析综述 ［J］. 计算机系统应用, 2013, 22(2): 1-4.
Wei W, Wang D D, Liu M, et al. Survey of human motion analysis ［J］. Computer Systems & Applications, 2013, 22(2):1-4.
［2］Elbrecht P, Palm K J. Precision of 3D body scanners［C］// 18th International Conference on Intelligent Engineering Systems. IEEE, 2014: 127-132.
［3］梅涛. 介电弹性体软传感器优化及在人体上肢运动测量中的应用 ［D］. 哈尔滨:哈尔滨工业大学, 2017.
［4］董秋雷, 吴毅红, 胡占义. 基于视频的实时自动人体高度测量 ［J］. 自动化学报, 2009, 35 (2): 137-144.
Dong Q L, Wu Y H, Hu Z Y. Video-based real-time automatic measurement for the height of human body［J］. Acta Automatica Sinica, 2009, 35 (2): 137-144.
［5］Jung J, Kim H, Yoon I, et al. Human height analysis using multiple uncalibrated cameras［C］ // IEEE Consumer Electronics Society. IEEE International Conference on Consumer Electronics. Las Vegas, NV,USA, 2016: 213-214.
［6］Jiang M X, Wang H, Qiu T. Video-based height measurements of multiple moving objects ［J］. KSII Transactions on Internet and Information Systems, 2014, 8(9): 3196-3210.
［7］李抵非, 陈赫, 冯志刚, 等. 一种双目立体视觉系统的校准方法 ［J］. 计量学报, 2018, 39(4): 485-489.
Li D F, Chen H, Feng Z G, et al. A calibration method of binocular vision system［J］. Acta Metrologica Sinica, 2018, 39(4): 485-489.
［8］赵莹莹, 曲昭伟, 江晟, 等. 基于视频的行人过街速度特性 ［J］. 吉林大学学报(工学版), 2013, 43(2): 298-303.
Zhao Y Y, Qu Z W, Jiang S, et al. Pedestrian crossing speed characteristic based on video detection［J］. Journal of Jilin University (Engineering and Technology Edition), 2013, 43(2): 298-303.
［9］吴凤和. 基于计算机视觉测量技术的图像轮廓提取方法研究 ［J］. 计量学报, 2007, 28 (1): 18-22.
Wu F H. A study on contour extraction method in computer vision measurement technology［J］. Acta Metrologica Sinica, 2007, 28(1): 18-22.
［10］Lv F, Zhao T, Nevatia R. Camera calibration from video of a walking human ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2006, 28(9): 1513-1518.
［11］Criminisi A, Reid I, Zisserman A. Single View Metrology ［J］. International Journal of Computer Vision, 2000, 40(2): 123-148.
［12］程淑红, 高许, 程树春, 等. 基于计算机视觉的运动车辆检测 ［J］. 计量学报, 2017, 38(3): 288-291.
Cheng S H, Gao X, Cheng S C, et al. Moving vehicle detection based on computer vision［J］. Acta Metrologica Sinica, 2017, 38(3): 288-291.
［13］孙忠胜, 王俊, 孙进平, 等. 单人行走运动参数估计方法 ［J］. 北京航空航天大学学报, 2016, 42(4): 789-795.
Sun Z S, Wang J, Sun J P, et al. Motion parameters estimation method of single walking human ［J］. Journal of Beijing University of Aeronautics and Astronautics, 2016, 42(4): 789-795.
2019, Vol. 40 
