Abstract
We have proposed novel techniques using region based segmentation for resolving complex occlusions of objects as they are tracked by a single, static video camera. The data is low resolution and noisy. The method improves the segmentation of moving objects obtained from any background subtraction method and then tracks objects continuously before, during and after occlusions. The method has been tested with real time complex occlusions of multiple people using a novel testing criterion that we have developed after critically analyzing all the possible scenarios of occlusions. The method has been proved to be successful with most of the testing criteria.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications|Proc. 3rd IASTED Int. Conf. Signal Process. Pattern Recogn. Applicat. |
| Pages | 140-146 |
| Number of pages | 6 |
| Volume | 2006 |
| Publication status | Published - 2006 |
| Event | 3rd IASTED International Conference on Signal Processing, Pattern Recognition, and Applications - Innsbruck Duration: 1 Jul 2006 → … |
Conference
| Conference | 3rd IASTED International Conference on Signal Processing, Pattern Recognition, and Applications |
|---|---|
| City | Innsbruck |
| Period | 1/07/06 → … |
Keywords
- 3D-histograms
- Occlusions
- Segmentations
- Testing criteria
- Tracking