Integration of computer vision in a testing facility for prosthetic joint inspection and performance assessment

Glanville, Stephen; Chong, Jun Jie; Jafari, Aghil; Etoundi, Appolinaire

Abstract

This paper presents a computer vision system within a test facility for prosthetic joint inspection and performance by analysing range of motion data. Testing with individuals rather than using a testing facility can cause issues, such as irritation, pain and rubbing of the residual limb. The use of a computer vision system within a testing facility can assess the performance of a prosthetic device by capturing video, images, and calculating angle data for range of motion. Other benefits of using a testing facility include reduced testing times and energy for the end user and the ability to test rigorously. It can be concluded that the computer vision system can capturing accurate angle data for use within the testing facility. However, a discrepancy between 0.55 and 0.6 degrees was found when comparing the angle of the knee joint in the computer aided design model to the built environment. Further development of the code and the use of masking with colour coded targets would result in more accurate angle calculation and automation.

Citation

Glanville, S., Chong, J. J., Jafari, A., & Etoundi, A. (2022). Integration of computer vision in a testing facility for prosthetic joint inspection and performance assessment. In 2021 24th International Conference on Mechatronics Technology (ICMT)https://doi.org/10.1109/icmt53429.2021.9687237

• Conference Name: 2021 24th International Conference on Mechatronics Technology (ICMT)
• Conference Location: Singapore
• Start Date: Dec 18, 2021
• End Date: Dec 22, 2021
• Online Publication Date: Feb 1, 2022
• Publication Date: Feb 1, 2022
• Deposit Date: Mar 9, 2022
• Publisher: Institute of Electrical and Electronics Engineers (IEEE)
• Book Title: 2021 24th International Conference on Mechatronics Technology (ICMT)
• ISBN: 9781665424608
• DOI: https://doi.org/10.1109/icmt53429.2021.9687237
• Public URL: https://uwe-repository.worktribe.com/output/9187163