Robust estimation of foreground in surveillance videos by sparse error estimation.
Mert DikmenThomas S. HuangPublished in: ICPR (2008)
Keyphrases
- robust estimation
- error estimation
- surveillance videos
- video surveillance
- model selection
- video data
- object classification
- object tracking
- least squares
- multi camera
- background subtraction
- surveillance system
- event detection
- moving objects
- video sequences
- image sequences
- motion field
- sparse representation
- real time
- video analysis
- generalization error
- compressed domain
- human activities
- image segmentation
- appearance model
- learning algorithm