Spark based distributed classification of spatial-spectral hyperspectral images

Remote sensing community recently turned its eye on Hyperspectral image classification. This research presents a distributed framework for supervised classification of Hyperspectral images using 2D?convolution based spatial and spectral information i.e., an HSI classification method based on fusion of both spatial and spectral information. The proposed procedure consists of two steps. The first step involves, feature selection method based on ANOVA F-test. The selected features are implemented on machine learning algorithm in distributed mode using Spark. For enhancing the classification accuracy, we considered 2D-convolution operation into account for spatial information. In the second step, the developed method is evaluated using benchmark datasets of hyperspectral images. The proposed approach yields better results in terms of classification accuracy when compared to other approaches. It also saves time the execution.