Edge Guided Structure Extraction for Hyperspectral Image Classification

In this paper, a novel edge guided structure extraction method is proposed for hyperspectral images classification, which consists of the following steps: First, the spectral dimension of the hyperspectral image is reduced with an averaging-based method. Then, the structural features is extracted by an extended relative total variation (ERTV) inspired by a learned edge probability map which serves as one of the major cues in the structure extraction process. Finally, the extracted structural features are fed into SVM for classification. Experimental results on two publicly available hyperspectral data sets demonstrate the competitive performance over several state-of-the-art classification approaches.