Fast measurement of human body posture based on three-dimensional optical information

Limei Song; Haozhen Huang; Yang Chen; Xinjun Zhu; Yangang Yang; Qinghua Guo

doi:10.3788/IRLA20200079

Journals >Infrared and Laser Engineering >Volume 49 >Issue 6 >Page 20200079 > Article

Infrared and Laser Engineering
Vol. 49, Issue 6, 20200079 (2020)

Fast measurement of human body posture based on three-dimensional optical information

Limei Song¹, Haozhen Huang¹, Yang Chen¹, Xinjun Zhu¹, Yangang Yang², and Qinghua Guo¹

Author Affiliations

¹天津工业大学天津市电工电能新技术重点实验室，天津 300387

²天津职业技术师范大学机械工程学院，天津 300222

show less

DOI: 10.3788/IRLA20200079 Cite this Article

Limei Song, Haozhen Huang, Yang Chen, Xinjun Zhu, Yangang Yang, Qinghua Guo. Fast measurement of human body posture based on three-dimensional optical information[J]. Infrared and Laser Engineering, 2020, 49(6): 20200079 Copy Citation Text

EndNote(RIS)

BibTex

Plain Text

show less

Fig. 1. Target data (a) distribution of the markers (b) angle calculation diagram

Download full size | View in the Article

Fig. 2. Binocular 3D human body scanning system

Download full size | View in the Article

Fig. 3. Calibration balls (a) calibration balls used in system measurement accuracy (b) schematic diagram of the spacing of the calibration balls

Download full size | View in the Article

Fig. 4. Method comparison (a) traditional TWPSP body scanning schematic (b) improved TWPSP body scanning schematic (c) fitting result of traditional method (d) result of the method in this paper (e) traditional TWPSP method reconstruction chart (f) improved TWPSP method reconstruction chart

Download full size | View in the Article

Fig. 5. Comparison of the average values of the three groups of poses

Download full size | View in the Article

Key dimensions		Traditional TWPSP method	Method in this paper
/mm	1	49.853	49.970
	2	50.053	50.038
	3	49.979	50.000
	4	49.956	50.009
	5	50.014	50.024
	Average	49.971	49.998
	RMSE	0.044	0.019
/mm	1	25.096	25.072
	2	24.975	25.019
	3	25.048	24.963
	4	24.981	25.051
	5	2525.077	24.949
	Average	25.061	25.028
	RMSE	0.053	0.037
/mm	1	55.654	55.924
	2	55.643	55.934
	3	55.572	55.910
	4	55.632	55.881
	5	55.610	55.875
	Average	55.940	55.895
	RMSE	0.055	0.035

Table 1. First five measurements of calibration balls

View in the Article

	\begin{document}${\hat d_1}$\end{document} /mm	\begin{document}${\hat d_{\rm{2}}}$\end{document} /mm	\begin{document}${\hat d_{\rm{3}}}$\end{document} /mm	\begin{document}$MA{E_{\rm{1}}}$\end{document} /mm	\begin{document}$MA{E_{\rm{2}}}$\end{document} /mm	\begin{document}$MA{E_{\rm{3}}}$\end{document} /mm
Traditional TWPSP method	19.949	19.961	19.978	0.048	0.035	0.039
Method in this paper	19.979	19.978	19.988	0.028	0.018	0.019

Table 2. Data comparison table of fitting results

View in the Article

Equipment	Point cloud number	Point cloud density	Accuracy/mm	Reconstruction time/s
Artec scanner	795 785	0.143	0.1	7
Sense scanner	121 717	0.001	0.9	5
Traditional TWPSP method system	236 540	0.016	0.055	2
Scanning system of this paper	296 540	0.016	0.03	1

Table 3. Point cloud data analysis of different sensors

View in the Article

	Average degree of No.1 car	Average degree of No.2 car	Average degree of No.3 car	RMSE of No.1 car	RMSE of No.2 car	RMSE of No.3 car
Angle A1	25.102	26.410	23.605	1.291	1.454	1.249
Angle A2	98.491	104.70 5	98.370	5.230	5.775	5.081
Angle A3	60.806	64.313	65.654	3.426	3.804	3.117
Angle A4	101.501	93.416	101.328	5.966	5.320	5.313
Angle A5	35.543	30.677	35.385	3.926	3.165	3.381
Angle A6	127.434	133.667	138.095	2.117	2.084	2.879
Angle A7	154.080	141.380	144.400	8.344	8.661	8.250
Angle A8	41.647	32.875	43.192	6.775	6.625	6.291
Angle A9	13.935	9.661	14.399	3.864	3.297	3.958

Table 4. Angles of key points (all values are reported as degree)

Download Citation

Save the article for my favorites

Paper Information