Segmentation of hyperspectral images using local covariance matrices in eigenspace.
Ugur ErgulGökhan BilginPublished in: SIU (2013)
Keyphrases
- covariance matrices
- hyperspectral images
- covariance matrix
- hyperspectral
- maximum likelihood
- gaussian mixture model
- gaussian distribution
- distance measure
- segmentation algorithm
- remote sensing
- segmentation method
- image segmentation
- vector space
- gaussian mixture
- image analysis
- region growing
- pixel classification
- multispectral
- principal component analysis
- feature vectors
- linear classifiers
- multispectral images
- multiscale
- edge detection
- infrared
- closed form
- upper bound
- image data
- similarity measure