Precision analysis of kinect to measure the motion of the upper body and calibration based on deep learning.

Saved in:

Bibliographic Details
Title:	Precision analysis of kinect to measure the motion of the upper body and calibration based on deep learning.
Authors:	Kyeong, Seulki¹ (AUTHOR), Kim, Yundong² (AUTHOR), Abunya, Philip² (AUTHOR), Kang, Bong-Soo² (AUTHOR) bskang@hnu.kr
Source:	Journal of Mechanical Science & Technology. Jun2025, Vol. 39 Issue 6, p3539-3545. 7p.
Subjects:	Motion capture (Cinematography), Measurement errors, Abduction (Kinesiology), Arm exercises, Deep learning, Shoulder exercises
Abstract:	This study analyzed the characteristics of the Kinect sensor developed for video games as a motion capture device. In addition, a calibration method was developed using deep learning technique to improve the precision of the Kinect while maintaining its convenience: no marker, low cost, easy-to-install, etc. When the proposed calibration scheme was applied to four types of arm exercises for three subjects, measurement errors were reduced by 68 % for shoulder movements and 43 % for elbow movements. Also, the measurement precision for shoulder flexion and abduction was improved up to 4 degrees, which sufficiently satisfied the requirement of rehabilitation exercises for patients with frozen shoulder. [ABSTRACT FROM AUTHOR]
	Copyright of Journal of Mechanical Science & Technology is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	This study analyzed the characteristics of the Kinect sensor developed for video games as a motion capture device. In addition, a calibration method was developed using deep learning technique to improve the precision of the Kinect while maintaining its convenience: no marker, low cost, easy-to-install, etc. When the proposed calibration scheme was applied to four types of arm exercises for three subjects, measurement errors were reduced by 68 % for shoulder movements and 43 % for elbow movements. Also, the measurement precision for shoulder flexion and abduction was improved up to 4 degrees, which sufficiently satisfied the requirement of rehabilitation exercises for patients with frozen shoulder. [ABSTRACT FROM AUTHOR]
ISSN:	1738494X
DOI:	10.1007/s12206-025-0545-2