A Novel Hierarchical Semisupervised SVM for Classification of Hyperspectral Images

Abstract

This letter presents a novel hierarchical semisupervised support vector machine (SVM) for classification of hyperspectral images. The method exploits the wealth of unlabeled samples by means of their cluster features. The method learns a suitable framework for classifying cluster features by a semisupervised SVM and thus makes use of advantages of clustering and classification. Experimental results demonstrate that the proposed classification method is effective for hyperspectral image classification when a few labeled samples are available. Another advantage of the proposed method is that the hierarchical structure can simultaneously take clustering and classification information into consideration.