University of Twente Student Theses
A Novel Approach Using Smoothing to Detect Errors in OpenPose Estimated Running Data
Krabbenborg, J. (2024) A Novel Approach Using Smoothing to Detect Errors in OpenPose Estimated Running Data.
|
PDF
4MB |
| Abstract: | OpenPose is a real-time, multi-person key point detection library that uses deep learning to identify and locate human body key points from images or videos and this can be used to calculate gait parameters. In this pose estimated data misdetections occur. However, the estimated data should be accurate to give correct advice to athletes and prevent injuries. In this research, a novel approach is developed to detect these errors using the moving average filter on OpenPose estimated running data. The data used in this research comprises datasets sourced from videos featuring individual runners from the sagittal plane. Performance metrics were calculated for two datasets using this approach. On the datasets the approach had an accuracy of 98% and 99% which means the approach has a very high ratio of correct predictions across both datasets. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Business & IT BSc (56066) |
| Link to this item: | https://purl.utwente.nl/essays/98165 |
| 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