Unsupervised Classification in Hyperspectral Imagery with Nonlocal Total Variation and Primal-Dual Hybrid Gradient Algorithm.
Wei ZhuVictoria ChayesAlexandre TiardStephanie SánchezDevin DahlbergDa KuangAndrea L. BertozziStanley J. OsherDominique ZossoPublished in: CoRR (2016)
Keyphrases
- primal dual
- total variation
- dual formulation
- unsupervised classification
- convex optimization
- augmented lagrangian
- image denoising
- linear programming
- denoising
- dynamic programming
- total variation regularization
- expectation maximization
- k means
- np hard
- computational complexity
- image restoration
- data clustering
- convergence rate
- approximation algorithms
- principal component analysis
- linear program
- similarity measure
- hyperspectral images
- linear programming problems
- natural images
- image segmentation
- pairwise
- optimal solution
- objective function
- reinforcement learning
- multiscale
- image sequences