Bibliometrics on the use of remote sensing and machine learning in crop classification

Abstract

This study presents a bibliometric analysis of global research on crop classification using remote sensing and machine learning (ML), a field critical to advancing precision agriculture. A systematic search in Scopus identified 2,122 peer-reviewed articles published between 2014 and 2023. The analysis employed VOSviewer and the Bibliometrix package in R to assess publication trends, citation impact, and keyword co-occurrence networks. Results reveal a marked increase in scientific production after 2017, coinciding with the availability of high-resolution satellite imagery and the adoption of deep learning algorithms, particularly convolutional neural networks (CNNs). China emerged as the leading contributor, followed by the United States and India, reflecting strong investments in agricultural modernization and remote sensing infrastructure. Thematic mapping highlights both traditional research areas, such as vegetation indices and land cover classification, and emerging themes, including AI-supported algorithms and food security. Despite this growth, disparities persist, with most countries contributing fewer than 100 publications, underscoring the need to promote participation in underrepresented regions. Findings demonstrate the field’s rapid evolution, emphasize the integration of AI-driven methods in crop monitoring, and suggest future directions combining remote sensing, ML, and internet of things (IoT) technologies to address global challenges in food security and sustainable agricultural management.