Improving superpixel-based image segmentation by incorporating color covariance matrix manifolds.
Xianbin GuJeremiah D. DengMartin K. PurvisPublished in: ICIP (2014)
Keyphrases
- normalized cut
- covariance matrix
- image segmentation
- covariance matrices
- image segmentation algorithm
- graph cuts
- image regions
- principal component analysis
- eigenvalues and eigenvectors
- mahalanobis distance
- riemannian manifolds
- sample size
- positive definite
- multiscale
- principal components
- multivariate gaussian
- high dimensional
- color space
- segmentation algorithm
- active contours
- correlation matrix
- manifold learning
- segmentation method
- eigendecomposition
- objective function
- geometrical interpretation
- image processing
- superpixels
- distance measure
- lie group
- feature extraction
- symmetric matrix