Hyperspectral Image Classification with Limited Labeled Training Samples Using Enhanced Ensemble Learning and Conditional Random Fields

Classification of hyperspectral imagery using few labelled samples is a challenging problem considering the high dimensionality of hyperspectral imagery. Classifiers trained on limited samples with abundant spectral bands tend to overfit, leading to weak generalization capability. Therefore, it is crucial to develop classification approaches that are capable of reducing model variance, while at the same time learning general structure from the training samples in an efficient manner. To address this problem, we have developed an enhanced ensemble method called multiclass boosted rotation forest (MBRF), which combines the rotation forest algorithm and a multiclass AdaBoost algorithm. The benefit of the combination can be explained by bias-variance analysis, especially in the situation of inadequate training samples and high dimensionality. Further, MBRF innately produces posterior probabilities inherited from AdaBoost, which are served as the unary potentials of the conditional random field (CRF) model to incorporate spatial context information. Experimental results show that the classification accuracy of MBRF as well as its integration with CRF consistently outperforms the other referenced state-of-the-art classification methods when limited labeled samples are available for training.