Quantitative Analysis of Different Land-use and Land-cover Classifiers using Hyperspectral Dataset

Remote sensing plays a vital role in analyzing land-use and land-cover (LULC) areas using hyperspectral imagery which was not feasible with multispectral imagery. The analysis of LULC is an important factor for environmental application. Classification is one of the methods used for the categorization of different types of class cateogries. The performance evaluation of various classifiers using hyperspectral dataset is essential to be performed for the categorization of various classes. Therefore, the main aim of this paper is to analyze and implement the various supervised classifiers such as random forest (RF), neural network (NN), and minimum distance classifier (MDC). The hyperspectral dataset has been used for these classifiers over a part of Haryana and Uttar Pradesh states, India. The results have shown that NN (88%) algorithm has achieved higher accuracy than other algorithms. These experimental outcomes highlights the potential of NN in handling complex problems and useful in the mapping of changes over LULC using hyperspectral imagery. This study is beneficial in applications such as crop monitoring, mapping seasonal variations, and snow detection.