Unsupervised Classification in Hyperspectral Imagery with Nonlocal Total Variation and Primal-Dual Hybrid Gradient Algorithm
This work addresses hyperspectral image analysis, offering an incremental improvement in unsupervised clustering for remote sensing applications.
The paper tackles unsupervised classification of hyperspectral images by proposing a graph-based nonlocal total variation method solved with a primal-dual hybrid gradient algorithm, achieving results that outperform standard methods like spherical K-means and NMF on synthetic and real-world data.
In this paper, a graph-based nonlocal total variation method (NLTV) is proposed for unsupervised classification of hyperspectral images (HSI). The variational problem is solved by the primal-dual hybrid gradient (PDHG) algorithm. By squaring the labeling function and using a stable simplex clustering routine, an unsupervised clustering method with random initialization can be implemented. The effectiveness of this proposed algorithm is illustrated on both synthetic and real-world HSI, and numerical results show that the proposed algorithm outperforms other standard unsupervised clustering methods such as spherical K-means, nonnegative matrix factorization (NMF), and the graph-based Merriman-Bence-Osher (MBO) scheme.