Geodesic active regions for segmentation and tracking of human gestures in sign language videos.
Olga DiamantiPetros MaragosPublished in: ICIP (2008)
Keyphrases
- sign language
- sign recognition
- hand tracking
- geodesic active
- sign language recognition
- video scene
- hand motion
- gesture recognition
- level set method
- hand gestures
- level set
- motion segmentation
- image segmentation
- foreground regions
- segmentation algorithm
- american sign language
- human activities
- segmentation method
- complex background
- video sequences
- object tracking
- edge detection
- shape prior
- dynamic scenes
- medical images
- human computer interaction
- particle filter
- video data
- input image
- object segmentation
- appearance model
- energy function
- motion analysis
- visual tracking
- spatio temporal
- active contours
- video images
- background subtraction
- computer vision
- moving objects
- multiscale
- image sequences
- image processing