A Novel Regularized Nonnegative Matrix Factorization for Spectral-Spatial Dimension Reduction of Hyperspectral Imagery.
Wei YanBob ZhangZuyuan YangPublished in: IEEE Access (2018)
Keyphrases
- hyperspectral imagery
- dimension reduction
- nonnegative matrix factorization
- principal component analysis
- spectral signatures
- least squares
- hyperspectral images
- hyperspectral
- multispectral
- singular value decomposition
- negative matrix factorization
- target detection
- feature extraction
- remote sensing
- matrix factorization
- random projections
- spatial resolution
- data representation
- infrared
- linear discriminant analysis
- dimensionality reduction
- feature space
- high dimensional data
- high dimensional
- cluster analysis
- high dimensional feature space
- low dimensional
- face recognition
- document clustering
- high dimensionality
- training data
- spectral clustering
- neural network
- objective function
- image processing
- multispectral images
- feature selection
- clustering algorithm
- unsupervised learning
- expectation maximization
- natural language processing
- optical flow