Dimensionality Reduction of Hyperspectral Images Based on the Linear Mixture Model and Dimensionality Estimation

摘要

In this paper, we propose a dimensionality reduction technique, which is based on the principal component analysisof homogenous spatial regions of hyperspectral images. In the proposed technique, we rely on the linear mixture modeland use a dimensionality estimation procedure to split an image into homogenous regions. The experiments carried outusing well-known hyperspectral image scenes show that the proposed technique allows obtaining compactrepresentations of image regions in reduced spectral subspaces and can be considered as a segmentation technique.