Application of convolutional neural network-based 3D posture estimation in behavioral analysis of construction workers

doi:10.16265/j.cnki.issn1003-3033.2019.07.011

Abstract

Abstract: In order to facilitate an automated behavioral analysis of construction workers, CNN was applied for 3D human pose estimation on sequential images. Considering the complicated site environment and dynamic operator behaviors, the data set of construction workers' postures was developed and the performance of the proposed algorithm was analyzed from both qualitative and quantitative aspects. Furthermore, the derived 3D postures in the video were used to drive the biomechanical model for more detailed and quantitative analysis of worker behavior. The results show that the proposed 3D pose estimation method is accurate and robust, and that combined with biomechanical model, more detailed analysis and evaluation of workers' behavior can be achieved.

Key words: 3D posture estimation, convolutional neural network (CNN), behavioral analysis, field test, construction safety

CLC Number:

X912.9

XIONG Ruoxin, SONG Yuanbin, WANG Yuxuan, DUAN Yanjuan. Application of convolutional neural network-based 3D posture estimation in behavioral analysis of construction workers[J]. China Safety Science Journal, 2019, 29(7): 64-69.

References

[1] 郭红领, 刘文平, 张伟胜. 集成BIM和PT的工人不安全行为预警系统研究[J]. 中国安全科学学报, 2014, 24(4): 104-109.
GUO Hongling, LIU Wenping, ZHANG Weisheng. A BIM-PT-integrated warning system for on-site workers' unsafe behavior [J]. China Safety Science Journal, 2014, 24(4): 104-109.
[2] PEDDI A, HUAN L, BAI Yong, et al. Development of human pose analyzing algorithms for the determination of construction productivity in real-time[C]. Construction Research Congress, 2009: 11-20.
[3] 晋良海, 张荣坤, 郑霞忠,等. 隧洞钻孔作业工效及低位作业姿势研究[J]. 中国安全科学学报, 2016, 26(4): 66-71.
JIN Lianghai, ZHANG Rongkun, ZHENG Xiazhong, et al. Research on simulation of drilling work posture and optimization of lowing work posture for tunnel [J]. China Safety Science Journal, 2016, 26(4): 66-71.
[4] 佟瑞鹏, 陈策, 刘思路,等. 面向行为安全的泛场景数据理论与应用研究[J]. 中国安全科学学报, 2017, 27(2): 1-6.
TONG Ruipeng, CHEN Ce, LIU Silu, et al. Research on theory and application of pan-scene data for behavioral safety [J]. China Safety Science Journal, 2017, 27(2): 1-6.
[5] 陈大伟, 田翰之, 张江石. 基于行为安全的建筑事故预防量化方法与实证研究[J]. 中国安全科学学报, 2010, 20(7): 96-102.
CHEN Dawei, TIAN Hanzhi, ZHANG Jiangshi. A quantitative method for construction accidents prevention based on behavior safety and its empirical study [J]. China Safety Science Journal, 2010, 20(4): 96-102.
[6] 韩豫, 张泾杰, 孙昊,等. 基于图像识别的建筑工人智能安全检查系统设计与实现[J]. 中国安全生产科学技术, 2016, 12(10): 142-148.
HAN Yu, ZHANG Jingjie, SUN Hao, et al. Design and implementation of intelligent safety inspection system for construction workers based on image recognition [J]. Journal of Safety Science and Technology, 2016, 12(10): 142-148.
[7] TOSHEV A, SZEGEDY C. DeepPose: human pose estimation via deep neural networks [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2014: 1 653-1 660.
[8] WEI S, RAMAKRISHNA V, KANADE T, et al. Convolutional Pose Machines [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2016: 4 724-4 732.
[9] NEWELL A, YANG Kaiyu, DENG Jiang. Stacked hourglass networks for human pose estimation [C]. European Conference on Computer Vision, 2016: 483-499.
[10] 熊若鑫, 宋元斌, 王宇轩. 基于DNN的作业姿势评估方法及应用[J]. 中国安全科学学报, 2018, 28(5): 105-110.
XIONG Ruoxin, SONG Yuanbin, WANG Yuxuan. Deep neural network-based posture assessment method and its application [J]. China Safety Science Journal, 2018, 28(5): 105-110.
[11] IQBAL U, DOERING A, YASIN H, et al. A dual-source approach for 3d human pose estimation from single images [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2016: 4 948-4 956.
[12] 尹燕芳, 孙农亮, 刘明,等. 基于BSCPs-RF的人体关节点行为识别与预测[J]. 机器人, 2017, 39(6): 795-802.
YIN Yanfang, SUN Nongliang, LIU Ming, et al. Action recognition and prediction of human skeleton based on BSCPs-RF [J]. Robot, 2017, 39(6): 795-802.
[13] SAPP B, TASKAR B. MODEC: multimodal decomposable models for human pose estimation [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2013: 3 674-3 681.
[14] JOHNSON S, EVERINGHAM M. Clustered pose and nonlinear appearance models for human pose estimation [C]. Proceedings of the British Machine Vision Conference, 2010: 12.1-12.11.
[15] ANDRILUKA M, PISHCHULIN L, GEHLER P, et al. 2D human pose estimation: new benchmark and state of the art analysis [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2014: 3 686-3 693.
[16] IONESCU C, PAPAVA D, OLARU V, et al. Human3.6M: large scale datasets and predictive methods for 3d human sensing in natural environments[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(7): 1 325-1 339.
[17] JOO H, LIU Hao, TAN Lei, et al. Panoptic studio: a massively multiview system for social motion capture [C]. IEEE International Conference on Computer Vision, 2015: 3 334-3 342.
[18] HIGNETT S, MCATAMNEY L. Rapid entire body assessment (REBA) [J]. Applied Ergonomics, 2000, 31(2): 201-205.
[19] 纪仲秋, 李旭龙, 姜桂萍,等. AnyBody仿真和验证的太极拳训练中下肢生物力学特征[J]. 中国康复医学杂志, 2014, 29(9): 799-805.
JI Zhongqiu, LI Xulong, JIANG Guiping, et al. A biomechanical research of lower limbs in taichiquan with simulation and modeling system [J]. Chinese Journal of Rehabilitation Medicine, 2014, 29(9): 799-805.
[20] National Institute for Occupational Safety. Work practices guide for manual lifting [R], 1981.