High-speed object tracking in ordinary surroundings based on temporally evaluated optical flow.
Ryuzo OkadaJunji OakiDaisuke YamamotoNobuhiro KondohHiroshi MiyazakiKoki UesugiJiro AmemiyaKenji ShirakawaAtsushi KunimatsuPublished in: IROS (2003)
Keyphrases
- object tracking
- optical flow
- high speed
- image sequences
- moving objects
- spatio temporal
- particle filter
- computer vision
- video sequences
- video surveillance
- mean shift
- appearance model
- kalman filter
- video surveillance systems
- illumination variations
- multiple cameras
- robust object tracking
- particle filtering
- multi camera
- visual tracking
- motion estimation
- appearance modeling
- multiple object tracking
- optical flow estimation
- real time
- tracking accuracy
- partial occlusion
- motion model
- sparse representation
- motion segmentation
- object location
- camera motion
- consistent labeling
- image classification
- moving camera
- flow field
- ego motion
- d scene
- pattern recognition
- image processing
- visual object tracking
- multiple motion models