Multiple Object Tracking
Open Access
- Author:
- Cui, Aiyu
- Area of Honors:
- Computer Science
- Degree:
- Bachelor of Science
- Document Type:
- Thesis
- Thesis Supervisors:
- Dr. Robert Collins, Thesis Supervisor
Danfeng Zhang, Thesis Honors Advisor - Keywords:
- Computer Vision
Tracking
Motion Capture
Taiji - Abstract:
- An important task in computer vision is tracking objects precisely to enable further data analysis. However, in reality, we can not guarantee that the tracking results out of a motion capture (MoCap) system are always accurate, due to lack of cameras numbers, limitation of camera coverage and so on. Unfortunately, manually fixing these errors can take more than 10 hours for a five-minute MoCap. Therefore, we are motivated to develop a method to fix imperfect motion capture data in terms of tracking correctness and gap filling. This thesis presents an innovative tracking algorithm to track objects formed by multiple trajectories with geometric correlations. In this thesis, we use a human skeleton formed by 39 markers (i.e. 39 trajectories) to illustrate how our algorithm can apply geometric constraints and select results among multiple results. We extend our algorithm to track multiple subjects (e.g. dual players and a human with a sword).
Accessible Version in Progress
We're generating an accessible version of this file to meet ADA Title II requirements. This process may take up to one hour. Please return later to access the accessible copy once it's ready.
You can still download the current version by clicking "OK".
What's happening:
An accessible PDF is being generated using Adobe with AI used to generate alternative text (alt text) for images in the PDF.