Exploiting multiple cues in motion segmentation based on background subtraction.
Ivan Huerta CasadoAriel AmatoF. Xavier RocaJordi GonzàlezPublished in: Neurocomputing (2013)
Keyphrases
- feature selection
- motion segmentation
- background subtraction
- moving objects
- image sequences
- model selection
- dynamic scenes
- video surveillance
- background model
- object tracking
- foreground objects
- video sequences
- motion detection
- surveillance system
- human detection
- moving camera
- foreground regions
- optical flow
- missing data
- motion model
- flow field
- motion estimation
- foreground and background
- camera motion
- object detection
- motion field
- gaussian mixture model
- appearance model
- spectral clustering
- active contours
- clustering algorithm
- computer vision