Pubescence Color Classification in Soybean Breeding Using Aerial Images and the Random Forest Machine Learning Algorithm

These data comprise a time series of aerial images taken from a drone and field notes of pubescence color from plant rows of the University of Illinois at Urbana-Champaign soybean breeding program. With this information and using the Random Forest algorithm, the fitted models classify gray, light tawny, and tawny pubescence. According to the year of data (2018-2020), the type of images (RGB or multispectral), and whether a single set or a time series of images is considered, each of these ten models can be tested using the code (.R files) and the model weights (.RData files).