Sparse Graph-Based Discriminant Analysis for Hyperspectral Imagery

• Published in: IEEE transactions on geoscience and remote sensing Vol. 52; no. 7; pp. 3872 - 3884
• Main Authors: Nam Hoai Ly, Qian Du, Fowler, James E.
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01.07.2014; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Applied geophysics; Classification; dimensionality reduction (DR); Discriminant analysis; Earth sciences; Earth, ocean, space; Eigenvalues and eigenfunctions; Exact sciences and technology; Graphs; hyperspectral imagery (HSI); Hyperspectral imaging; Imagery; Internal geophysics; Kernel; Kernels; Optimization; Principal component analysis; Projection; Reduction; Sparse matrices; sparse representation; Support vector machines; hyperspectral imagery (HSI); dimensionality reduction (DR); Classification; sparse representation; discriminant analysis; imagery; remote sensing; projection; classification; reduction; performances; quality
• ISSN: 0196-2892; 1558-0644
• DOI: 10.1109/TGRS.2013.2277251

Sparsity-preserving graph construction is investigated for the dimensionality reduction of hyperspectral imagery. In particular, a sparse graph-based discriminant analysis is proposed when labeled samples are available. By forcing the projection to be along the direction where a sample is clustered with within-class samples that best represented it, the discriminative power can be enhanced. The proposed method has no requirement on the number of labeled samples as in traditional linear discriminant analysis, and it can be solved by a simple generalized eigenproblem. The quality of the dimensionality reduction is evaluated by a support vector machine with a composite spatial-spectral kernel. Experimental results demonstrate that the proposed sparse graph-based discriminant analysis can yield superior classification performance with much lower dimensionality as compared to performance on the original data or on data transformed with other dimensionality-reduction approaches.