Semisupervised Band Clustering for Dimensionality Reduction of Hyperspectral Imagery

TLDR: The experimental results show that the proposed semisupervised band clustering algorithm can outperform other existing methods with lower computational cost.

Abstract: Band clustering is applied to dimensionality reduction of hyperspectral imagery. Different from unsupervised clustering using all the pixels or supervised clustering requiring labeled pixels, the proposed semisupervised band clustering needs class spectral signatures only. After clustering, a cluster selection step is applied to select clusters to be used in the following data analysis. Initial conditions and distance metrics are also investigated to improve the clustering performance. The experimental results show that the proposed algorithm can outperform other existing methods with lower computational cost.