Combining 3D Shape, Color, and Motion for Robust Anytime Tracking.
David HeldJesse LevinsonSebastian ThrunSilvio SavaresePublished in: Robotics: Science and Systems (2014)
Keyphrases
- robust tracking
- multiple cues
- tracking framework
- motion analysis
- motion model
- motion tracking
- image sequences
- particle filter
- motion segmentation
- visual tracking
- motion detection
- illumination change
- partial occlusion
- object tracking algorithm
- color images
- color based tracking
- object tracking
- fundamental problem in computer vision
- moving camera
- real time tracking
- deformable objects
- single image
- moving target
- articulated objects
- optical flow
- articulated body
- extended kalman filters
- dynamic background
- rigid objects
- temporal continuity
- motion field
- appearance model
- moving objects
- d objects
- color space
- beating heart
- long video sequences
- target object
- color model
- consecutive frames
- complex background
- three dimensional
- particle filtering
- kalman filter
- mean shift
- space time
- object motion
- motion patterns
- dynamic scenes
- body parts
- human motion
- video sequences
- wide area motion imagery