University of Twente Student Theses
The comparison of the accuracy of camera-based skeleton trackers during the stepping response for recovering balance in a home based environment.
Nieuwenhuys, W.W. (2022) The comparison of the accuracy of camera-based skeleton trackers during the stepping response for recovering balance in a home based environment.
|
PDF
4MB |
| Abstract: | For the 350 000 stroke survivors in the Netherlands, frequent falling is one of the most common medical complications. The HEROES Project aims to reduce fall-related healthcare utilization and help stroke patients stay independent. This is done by developing an exergaming system for at-home rehabilitation, which is focused on the stepping response of patients. A skeleton tracking algorithm is used to estimate the joint locations. Various skeleton tracking algorithms exist without clear information on which performs the best in an at-home environment. Two experiments have been performed to quantify the quality of skeleton tracking during a stepping exercise by three skeleton trackers: Cubemos, Nuitrack and Openpose. During one experiment, the three algorithms were compared to a golden standard, Qualisys. The second experiment compared the algorithms during the same exercise in a home-based environment, analysing the effect of ambient conditions and user properties Cubemos is the most robust skeleton tracker to changes in a home-based environment. For the best joint recognition, it is advised to turn on lights and wear monochromatic clothing. Even though steps in different directions could easily be identified, all skeleton trackers show a large error to Qualisys and are not deemed suitable for precise location estimations |
| Item Type: | Essay (Master) |
| Faculty: | ET: Engineering Technology |
| Subject: | 30 exact sciences in general, 44 medicine, 50 technical science in general |
| Programme: | Biomedical Engineering MSc (66226) |
| Link to this item: | https://purl.utwente.nl/essays/92178 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page