A New Foreground and Background Image Segmentation Method Based on a Convex Shape Prior and a Nonconvex Regularizer.
Yu HanXuyuan ZhangChen XuPublished in: IEEE Access (2022)
Keyphrases
- segmentation method
- shape prior
- foreground and background
- object segmentation
- active contours
- foreground objects
- image segmentation
- active contour model
- segmentation algorithm
- intensity information
- region growing
- superpixels
- energy function
- segmentation result
- watershed segmentation
- global optimization
- bayesian framework
- mean shift
- region merging
- motion segmentation
- graph cuts
- pixel level
- prior shape knowledge
- energy functional
- background subtraction
- deformable models
- markov random field
- image analysis