|
(See ISLA for the new IAS website. Most maintenance activities on this old website have stopped.)
Agression
 |
Image sequence analysis for security and surveillance |
Objective
In a joint project with TNO Physics and Electronics Laboratory we investigate the use of image processing on image sequences for security and surveillance application. We aim at making a system capable of object detection and tracking which:
- Is robust to varying global illumination.
- Can track objects deformable objects.
- Tracks objects through occlusion.
Research group members
ir. P.J. Withagen
prof.dr.ir. F.C.A. Groen
Funding
This research is funded by the Research School Safety and Security in Society (OMV) and TNO
"Kenniscentrum Intelligente Autonome Systemen".
Research Achievements
We are currently working on a tracking algorithm which can robustly track single persons under occlusion of objects and other people, using a single color-camera. Some achievements are:
- Automatic correction of global intensity changes due to camera auto-gain or changes in illumination.
- Background modelling and object detection using a probabilistic approach.
- Probabilistic integration of background modelling and object tracking.
More information
|
|
