Hyperspectral Unmixing Using Weighted L1/2 Sparse Total Variation Regularized and Volume Prior Constrained Nonnegative Matrix Factorization.
Kewen QuWenxing BaoXiangfei ShenPublished in: IGARSS (2019)
Keyphrases
- total variation
- nonnegative matrix factorization
- hyperspectral
- least squares
- image restoration
- denoising
- hyperspectral images
- remote sensing
- image denoising
- multispectral
- hyperspectral data
- infrared
- negative matrix factorization
- regularized least squares
- data representation
- convex optimization
- matrix factorization
- image data
- random projections
- sparse representation
- hyperspectral imagery
- image processing
- spectral clustering
- interior point
- objective function
- information content
- high dimensional
- document clustering
- principal component analysis
- original data
- satellite images
- regularization parameter
- text mining
- collaborative filtering
- high resolution
- multiresolution
- pairwise