Species classification using hyperspectral data and model evaluation

Species classification is an important task when it comes to the study of specific areas without disrupting the ecosystem. In this paper we chose Indian Pines as the test site to classify it into 16 classes of tree species and compare them with ground truth to obtain the accuracy, using various algorithms with machine learning process. The algorithms used are LDA and SVM. Spectral features of trees are obtained from hyperspectral data. Results show improvement in classification performance. We obtain an accuracy of 82 percent for LDA and 84 percent for SVM. Our aim is to indicate which algorithm works best for test site.