Automatic people segmentation with a template-driven graph cut.
Cyrille MigniotPascal BertolinoJean-Marc ChasseryPublished in: ICIP (2011)
Keyphrases
- graph cuts
- shape prior
- segmentation algorithm
- energy function
- object segmentation
- graph cut segmentation
- image segmentation
- interactive segmentation
- interactive image segmentation
- graph based image segmentation
- spatial coherence
- random walker
- energy minimization
- foreground background segmentation
- min cut
- markov random field
- chan vese
- robust segmentation
- segmentation accuracy
- belief propagation
- pixel labeling
- graph cut optimization
- globally optimal
- fully automatic
- geodesic active contours
- framework for image segmentation
- labeling problems
- mean shift
- segmentation method
- normalized cut
- region growing
- level set
- automatic segmentation
- deformable models
- segmentation result
- active contours
- multi label
- superpixels
- energy functional
- regularization term
- pairwise
- max flow
- prior information
- background subtraction
- medical imaging
- prior shape
- depth map
- loopy belief propagation
- user interaction
- medical images
- input image