Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Independently moving multi-camera system for cyclist motion capture
Thesis (MEng)--Stellenbosch University, 2024.
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
Stellenbosch : Stellenbosch University
2025
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867613910542057472 |
|---|---|
| access_status_str | Open Access |
| author | De Jongh, Mila |
| author2 | Schreve, Kristiaan |
| author_browse | De Jongh, Mila Schreve, Kristiaan |
| author_facet | Schreve, Kristiaan De Jongh, Mila |
| author_sort | De Jongh, Mila |
| collection | Thesis |
| dc_rights_str_mv | Stellenbosch University |
| description |
Thesis (MEng)--Stellenbosch University, 2024. |
| format | Thesis |
| id | oai:scholar.sun.ac.za:10019.1/131634 |
| institution | Stellenbosch University (South Africa) |
| language | English |
| last_indexed | 2026-06-10T12:43:39.397Z |
| license_str | Other — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from SUNScholar — Stellenbosch University Repository |
| publishDate | 2025 |
| publishDateRange | 2025 |
| publishDateSort | 2025 |
| publisher | Stellenbosch : Stellenbosch University |
| publisherStr | Stellenbosch : Stellenbosch University |
| record_format | dspace |
| source_str | SUNScholar — Stellenbosch University Repository |
| spelling | oai:scholar.sun.ac.za:10019.1/131634 Independently moving multi-camera system for cyclist motion capture De Jongh, Mila Schreve, Kristiaan Stellenbosch University. Faculty of Engineering. Dept. of Mechanical and Mechatronic Engineering. Human activity recognition Cycling -- Ability testing Digital cameras UCTD Thesis (MEng)--Stellenbosch University, 2024. ENGLISH ABSTRACT: A multi-camera motion capturing system using marker-based techniques are investigated to reliably capture the motion of a cyclist for performance improvement or rehabilitation purposes. The system uses cameras moving independently and unpredictably. The research investigates factors and algorithm limitations, such as noise in the image points, depth information, baseline distances between cameras, and multiple solutions to the fundamental matrix, that can influence the final results. Challenges regarding camera self-calibration, synchronization and marker detection and identification are highlighted. The widely used Vicon motion capture system, considered the gold standard, was used to validate the proposed system’s results, along with additional experiments to confirm its feasibility and reliability. The range of motion for different angles and a knee offset was measured and compared to similar measurements with the Vicon system. For some of these measurements, the average deviation between the two systems was 18 ° for the Hip-Knee-Ankle angle, 2.7 ° for the Elbow-Shoulder-Hip angle, and 3.1 mm for the knee offset relative to the hip-ankle line. It is important to note that a direct comparison with the Vicon system was not entirely possible due to differences in marker types and placements, as well as the Vicon system's use of a body model to determine joint centres, resulting in slightly different angles measured between the two systems. Recommendations for future work include enhancements in marker placement, camera distance management, and the extended use of AI body models. The findings offer valuable insights for advancing motion capture technology in sports and other fields. AFRIKAANSE OPSOMMING: Die ondersoek behels 'n multi-kamera bewegingsopname sisteem wat merkergebaseerde tegnieke gebruik om die beweging van 'n fietsryer betroubaar vas te lê vir prestasieverbetering of rehabilitasie doeleindes. Die stelsel gebruik kameras wat onafhanklik en onvoorspelbaar van mekaar beweeg. Die navorsing ondersoek faktore en algoritmebeperkings, soos geraas in die beeldpunte, diepte-inligting, basislynafstande tussen kameras, en veelvuldige oplossings vir die fundamentele matriks, wat die finale resultate kan beïnvloed. Uitdagings met betrekking tot kamera self-kalibrasie, sinchronisasie en merkeridentifikasie word bespreek. Die Vicon-bewegingsopname-stelsel, wat as die goue standaard beskou word, is gebruik om die voorgestelde stelsel se resultate te valideer, saam met bykomende eksperimente om die uitvoerbaarheid en betroubaarheid daarvan te bevestig. Die bewegingsomvang vir verskillende hoeke en 'n knie verplasing is gemeet en vergelyk met soortgelyke metings met die Vicon-stelsel. Vir sommige van hierdie metings tussen die twee sisteme was die gemiddelde afwyking 18° vir die Heup-Knie-Enkel hoek, 2.7° vir die Elmboog-Skouer-Heup hoek, en 3.1 mm vir die verplasing van die knie relatief tot die heup-enkel lyn. Dit is belangrik om te weet dat 'n direkte vergelyking met die Vicon-stelsel nie moontlik was nie as gevolg van verskille in merker tipes en plasing, sowel as die Vicon-stelsel se gebruik van 'n liggaamsmodel om gewrig middelpunte te bepaal. Dit het gelei tot effens verskillende hoeke wat gemeet is tussen die twee stelsels. Aanbevelings vir toekomstige werk sluit in verbeterings in merkerplasing, kamera basislynafstand beheer, en die uitgebreide gebruik van Kunsmatige Intelligensie liggaamsmodelle. Die bevindinge bied waardevolle insigte vir die bevordering van bewegingsopname-tegnologie in verskeie sportsoorte en ander velde. Masters 2025-02-03T05:54:53Z 2025-02-03T05:54:53Z 2024-12 Thesis https://scholar.sun.ac.za/handle/10019.1/131634 en Stellenbosch University xv, 120 pages : illustrations application/pdf Stellenbosch : Stellenbosch University |
| spellingShingle | Human activity recognition Cycling -- Ability testing Digital cameras UCTD De Jongh, Mila Independently moving multi-camera system for cyclist motion capture |
| title | Independently moving multi-camera system for cyclist motion capture |
| title_full | Independently moving multi-camera system for cyclist motion capture |
| title_fullStr | Independently moving multi-camera system for cyclist motion capture |
| title_full_unstemmed | Independently moving multi-camera system for cyclist motion capture |
| title_short | Independently moving multi-camera system for cyclist motion capture |
| title_sort | independently moving multi camera system for cyclist motion capture |
| topic | Human activity recognition Cycling -- Ability testing Digital cameras UCTD |
| url | https://scholar.sun.ac.za/handle/10019.1/131634 |
| work_keys_str_mv | AT dejonghmila independentlymovingmulticamerasystemforcyclistmotioncapture |