A comparative study of classifiers for analyzing agricultural land use

Abstract

This paper focuses on a comparative study of the accuracy of classifiers for aerial photographic land use to reduce staff burden in surveying areas of integrated agriculture. The widest algorithms are used for object classification in the field of geographic information systems consisting of Maximum Likelihood (ML), Support Vector Machine (SVM) and Random Forest (RF). These algorithms are benchmarked for the classification of agricultural land. An aerial photograph is captured by cameras mounted on Unmanned Aerial Vehicle (UAV) to record red band images and near-infrared images to calculate the Normalized Difference Vegetation Index (NDVI). As a result, the RF algorithm is the most accurate, equal to 93 percent.