Resolving occlusion ambiguity by combining Kalman tracking with feature tracking for image sequences.
Mark HeimbachKamak EbadiSally L. WoodPublished in: ACSSC (2017)
Keyphrases
- feature tracking
- image sequences
- motion analysis
- structure from motion
- optical flow
- visual motion
- long image sequences
- motion segmentation
- motion model
- motion parameters
- multi frame
- scene structure
- feature point tracking
- optical flow algorithm
- bundle adjustment
- camera motion
- feature detection
- motion estimation
- computer vision
- motion field
- kalman filtering
- camera parameters
- kalman filter
- object tracking
- video sequences
- moving camera
- computationally efficient
- motion patterns
- partial occlusion
- preprocessing
- particle filter
- ground plane
- filtering algorithm
- image motion
- three dimensional
- visual tracking
- face recognition
- feature points
- articulated objects
- model selection
- feature matching
- d scene
- motion detection
- closed form
- point correspondences
- background subtraction