A Comparison of Machine Learning Techniques Applied to Uav Data for Nitrogen Content Estimation.

The use of Unmanned Aerial Vehicles (UAV) for precision agriculture (PA) applications has been increasing over the last decade because of their potential to be a low-cost, accessible, and practical substitute for satellite and aircraft for acquiring high temporal and spatial resolution remotely sensed data. A rotary-wing UAV equipped with a multispectral sensor (Rededge, Micasense Systems) was used to acquire very high spatial resolution images of spring wheat plots. Then the performance of Support Vector Machine (SVM) and Relevance Vector Machine (RVM) approaches were assessed in estimating spring wheat nitrogen (N) content using UAV-acquired data. The experiment was conducted on spring wheat plots with five N levels (0, 84, 168, 252, and 336 kg N ha^-1) and four replications at five different locations in Idaho in 2016 and 2017. Validation of the methods was based on the cross-validation procedure and using three statistical indicators: R², RMSE and relative RMSE. The cross-validated results identified all models providing accurate estimates of canopy N content in spring wheat crop.