Enhancing agronomical attribute detection through RPAS-based image analysis - A review

The integration of Remotely Piloted Aircraft Systems (RPAS) into agronomical research has revolutionized the collection and analysis of spatial and temporal crop data. RPAS, equipped with high-resolution multispectral, thermal and visible cameras, provide a cost-effective and flexible alternative to traditional laboratory and manual methods. This review synthesizes recent advancements in RPAS-based image analysis for agronomic applications, with a focus on crop monitoring, weed detection, biomass estimation and yield prediction. A critical evaluation of the published information reveals that most studies utilize low-altitude flights with commercial drones integrated with sensors capable of capturing data with high spatial resolution. Image processing techniques, such as vegetation indices, machine learning algorithms and object-based image analysis, are commonly employed to extract biophysical and biochemical parameters. The review of the literature demonstrates a strong correlation between RPAS-derived metrics and ground-based measurements, validating their utility in precision agriculture. However, variability in sensor calibration, flight parameters and environmental conditions presents challenges to reproducibility and scalability. Overall, RPAS-based image analysis offers a promising avenue for enhancing data-driven decision-making in agriculture, contributing to more sustainable and efficient farming practices in the future.