Efficient classification of the hyperspectral images using deep learning

Classification techniques applicable to the hyperspectral images do not extract deep features from the hyperspectral image efficiently. In this work, a deep learning approach is proposed to extract the deep features, and these features are utilized to propose a novel framework for classification of the hyperspectral image. The framework uses LPP, DCNN and logistic regression. Data of a hyperspectral image is processed by LPP for dimensionality reduction as it contains a large number of dimensions. Afterward, a DCNN is constructed with Autoencoders which is then passed to the logistic regression for classification. Proposed framework is tested on Indian Pines and Salinas data sets. High accuracy is achieved using the proposed framework in comparison of existing machine learning models.