Activity Segmentation and Fish Tracking From Sonar Videos by Combining Artifacts Filtering and a Kalman Approach.
Julian WinklerSabah Badri-HoeherFatna BarkouchPublished in: IEEE Access (2023)
Keyphrases
- foreground background segmentation
- kalman filter
- filtering algorithm
- motion segmentation
- motion cues
- human activities
- video stabilization
- segmentation algorithm
- activity analysis
- image filtering
- video scene
- kalman filtering
- segmentation accuracy
- automatic initialization
- level set
- segmentation method
- video surveillance
- dynamic scenes
- real time
- surveillance videos
- video sequences
- moving camera
- multiple cues
- image segmentation
- video analysis
- multiscale
- mobile robot
- graph cuts
- group activities
- medical images
- video images
- particle filtering
- high quality
- motion estimation
- text detection
- target recognition
- complex background
- shape prior
- particle filter
- video segments
- video objects
- filtering method
- mean shift
- motion model
- video frames
- high resolution
- moving objects
- multiple objects