Kernel Grouped Multivariate Discriminant Analysis for Hyperspectral Image Classification

This paper proposes a grouping based technique of multivariate anal-ysis, and it is extended to nonlinear kernel based version for hyperspectral im-age classification. Grouped multivariate analysis methods are presented in the Euclidean space and dot products are replaced by kernels in Hilbert space for nonlinear dimension reduction and data visualization. We show that the pro-posed kernel analysis method greatly enhances the classification performance. Experiments on Classification are presented based on Indian Pine real dataset collected from the 224-dimensional AVIRIS hyperspectral sensor, and the per-formance of proposed approach is investigated. Results show that the Kernel Grouped Multivariate discriminant Analysis (KGMVA) method is generally ef-ficient to improve overall accuracy.