Object Tracking in Satellite Videos Based on a Multi-Frame Optical Flow Tracker.
Bo DuShihan CaiChen WuLiangpei ZhangDacheng TaoPublished in: CoRR (2018)
Keyphrases
- object tracking
- multi frame
- optical flow
- image sequences
- temporal coherence
- video sequences
- video surveillance
- moving camera
- motion estimation
- moving objects
- appearance model
- feature tracking
- computer vision
- video frames
- surveillance videos
- kalman filter
- motion analysis
- visual tracking
- motion detection
- particle filter
- camera motion
- optical flow estimation
- mean shift
- motion parameters
- flow field
- motion segmentation
- motion model
- dynamic scenes
- vector field
- motion field
- multi camera
- multiple cameras
- three dimensional
- motion trajectories
- background subtraction
- structure from motion
- feature points
- robust tracking
- tracking objects
- visual object tracking
- object tracking algorithm
- human motion
- d scene
- video data
- computationally efficient
- super resolution
- optic flow
- ego motion