An adaptive band selection algorithm for dimension reduction of hyperspectral images

Abstract

An adaptive band selection algorithm for dimension reduction of hyperspectral images is proposed. Considering the spatial correlation and spectral correlation, a selection rule, referring to spectral information and its correlation, is constructed for band selection. To test the efficiency of this algorithm, K-means algorithm for unsupervised classification was applied on images generated from the algorithm. The results showed that the proposed algorithm reduced the computation amount and improved the classification accuracy.