Band elimination of hyperspectral imagery using correlation of partitioned band images

In this article, an unsupervised band elimination method for hyperspectral imagery has been proposed which iteratively eliminates one band from the pair of most correlated neighboring bands depending on discriminating capability of the bands. Correlation between neighboring bands is calculated over partitioned band images. Capacitory discrimination is used to measure the discrimination capability of a band image. To demonstrate the effectiveness of the proposed method, results are compared with three state-of-the-art methods in terms of overall classification accuracy and Kappa coefficient. Results for the proposed methodology are found to be encouraging.