VHR satellite image segmentation based on topological unsupervised learning.
Nistor GrozavuNicoleta RogovschiGuénaël CabanesAndres Troya-GalvisPierre GançarskiPublished in: MVA (2015)
Keyphrases
- remote sensing
- unsupervised learning
- image segmentation
- change detection
- aerial images
- image analysis
- image processing
- remote sensing images
- expectation maximization
- medical imaging
- supervised learning
- multispectral
- graph cuts
- high resolution
- hyperspectral
- deep learning
- satellite images
- active contours
- data clustering
- level set method
- computer vision
- markov random field
- multiscale
- dimensionality reduction
- semi supervised
- method for image segmentation
- machine learning
- object recognition
- segmentation method
- deep architectures
- probabilistic relaxation
- contour detection
- pattern recognition
- object segmentation
- image registration
- deformable models
- region growing
- graph partitioning
- normalized cut
- topological properties
- topological features
- level set
- watershed transformation
- model selection
- gray level
- feature selection
- boundary detection
- image segmentation algorithm
- image segmentation algorithms