Clustering Motion for Real-Time Optical Flow Based Tracking.
Tobias SenstRubén Heras EvangelioIvo KellerThomas SikoraPublished in: AVSS (2012)
Keyphrases
- optical flow
- real time
- motion model
- motion segmentation
- feature tracking
- real time tracking
- image sequences
- motion field
- monocular camera
- lucas kanade
- color based tracking
- moving camera
- dense optical flow
- moving target
- consecutive frames
- motion estimation
- optical flow estimation
- visual odometry
- motion analysis
- motion parameters
- camera motion
- spectral clustering
- optical flow computation
- high frame rate
- detecting moving objects
- image motion
- optical flow algorithm
- active camera
- ego motion
- clustering algorithm
- flow field
- beating heart
- surveillance videos
- moving objects
- robust tracking
- motion detection
- optical flow fields
- motion tracking
- visual motion
- motion patterns
- particle filter
- moving observer
- structure from motion
- multiple moving objects
- d scene
- rigid body motion
- video sequences
- motion discontinuities
- visual tracking
- kalman filter
- background subtraction
- particle filtering
- smoothness constraint
- motion features
- appearance model
- object motion
- motion boundaries
- motion estimates
- scene flow