3D modeling for crop applications in precision agriculture: a review

Background Advances in data acquisition and information processing have strengthened the role of precision agriculture (PA) in enhancing agricultural productivity and sustainability. Among emerging technologies, 3D crop modeling offers detailed representations of agricultural scenes, supports crop phenotyping, and enables autonomous operations such as harvesting, fertilization, and pruning. However, despite the rapid development of hardware and software for 3D modeling, a comprehensive synthesis of the available approaches and their applications in PA remains limited. Purpose This review aims to examine the broad range of 3D modeling approaches used in PA, focusing on their methodological characteristics, practical applications, advantages, and limitations to support the selection of suitable solutions for different agricultural requirements. Methods A total of 168 publications, identified through a Scopus literature search covering studies published up to March 2026, were reviewed. The analysis examines 3D data types and sensing technologies, sensor platforms and application environments, and data processing methods, including data registration techniques and methods for insight extraction. Applications of 3D models to agricultural crops were also analyzed and classified into three main categories: crop phenotyping, machinery navigation, and field operations. Conclusions The review highlights the widespread adoption of LiDAR, RGB, and RGB-D sensors deployed on Uncrewed Aerial Vehicles, Uncrewed Ground Vehicles, and robotic platforms, alongside increasing use of AI-based processing methods. Challenges related to sensor robustness, multisensor interoperability, and large 3D datasets remain. Future research should focus on overcoming these limitations to facilitate the transition towards reliable, large-scale implementation of 3D modeling technologies in practical agricultural operations.