Motion and Appearance Contexts for Tracking and Re-Acquiring Targets in Aerial Videos.
Saad AliVladimir ReillyMubarak ShahPublished in: CVPR (2007)
Keyphrases
- motion cues
- appearance features
- moving target
- moving camera
- appearance model
- object motion
- input video
- motion model
- static images
- motion features
- motion analysis
- recognition of human actions
- successive frames
- appearance information
- dynamic scenes
- tracking multiple
- particle filter
- traffic scenes
- dynamic textures
- video sequences
- motion tracking
- observation model
- motion segmentation
- camera motion
- video frames
- robust tracking
- video stabilization
- long video sequences
- target object
- motion detection
- image sequences
- visual tracking
- visual motion
- object tracking
- multiple targets
- target tracking
- motion patterns
- optical flow
- articulated body
- surveillance videos
- appearance variations
- low frame rate
- space time
- periodic motion
- moving objects
- multiple target tracking
- articulated human motion
- action recognition
- video surveillance
- tracking framework
- wide area motion imagery
- video scene
- articulated objects
- key frames
- object appearance
- motion estimation
- motion field
- motion sequences
- dynamic background
- foreground regions
- multi view
- motion parameters
- human motion
- upper body
- foreground objects
- temporal continuity
- human activities
- kalman filter
- human pose
- feature points
- video data
- mean shift
- d scene
- motion capture
- human actions
- particle filtering
- video analysis