F. Mendiburu, P. Delicado, R. Morales, H. Loayza, R. Quiroz, E. Schrevens
Remote sensing from an unmanned aerial vehicle (UAV, or drone) in agriculture is one of the most appropriate ways to obtain information about crop growth in large areas. Additionally, drones are also able to provide detailed information in agricultural experimentation. Our aim is to propose a systematic methodology that allows transforming reflectance images obtained by drones in serial flights over a crop, into functional data. The experimental units (which each functional data is linked to) are individual plants (if they are identifiable from the drone) or small areas containing a few plants. Our claim is that reflectance functions capture the growth dynamics of plants at least as well as manual measurements taken in situ. They are also cheaper and cover the entire crop field. We illustrate our proposals with a real experiment developed at the International Potato Center in Peru with cassava, maize and sweet potato crops. The drone flew the field regularly, recording 38 images.
Palabras clave / Keywords: remote sensing, crop growth curves, GAM model
Programado
Sesión V03 Análisis de Datos Funcionales
1 de junio de 2018 16:00
Sala 2
