Kernel Fukunaga-Koontz Transform Subspaces for Classification of Hyperspectral Images With Small Sample Sizes.
Hamidullah BinolGökhan BilginSemih DinçAbdullah BalPublished in: IEEE Geosci. Remote. Sens. Lett. (2015)
Keyphrases
- fukunaga koontz
- hyperspectral images
- sample size
- small sample
- hyperspectral data
- hyperspectral
- model selection
- support vector
- feature space
- remote sensing
- multispectral
- hyperspectral imagery
- spectral bands
- small samples
- random sample
- feature extraction
- pattern recognition
- confidence intervals
- target detection
- multispectral images
- random sampling
- pixel classification
- machine learning
- high quality
- information content
- infrared
- kernel function
- low dimensional