Dense estimation and object-based segmentation of the optical flow with robust techniques.
Étienne MéminPatrick PérezPublished in: IEEE Trans. Image Process. (1998)
Keyphrases
- optical flow
- dense motion estimation
- optical flow estimation
- dense optical flow estimation
- optical flow field
- motion field estimation
- dense optical flow
- motion segmentation
- optical flow fields
- independently moving objects
- orientation estimation
- scene flow
- optical flow computation
- robust segmentation
- multiscale
- image sequences
- least squares
- flow field
- flow estimation
- motion estimation
- estimation error
- robust estimation
- smoothness constraint
- image segmentation
- structure tensor
- motion model
- motion field
- motion boundaries
- dense correspondence
- segmentation algorithm
- estimation of optical flow
- object segmentation
- image motion
- level set
- localization method
- velocity field
- feature tracking
- ellipse fitting
- segmentation method
- maximum likelihood criterion
- computer vision
- moving objects
- camera motion
- qualitative and quantitative evaluation
- vector field
- iris segmentation
- partial volume
- region growing
- shape prior
- multibody
- displacement field