Real-time kinematics applied at unmanned aerial vehicles positioning for orthophotography in precision agriculture

In recent years, Unmanned Aerial Vehicles (UAVs) have gained attention in the agricultural field, particularly in what is known as Precision Agriculture (PA). The use of these vehicles responds to the need for optimizing resources to have greater yields while reducing the negative impact on the environment. One of the main functions for UAVs in this field is the Remote Sensing (RS) of the crops. A UAV can fly over a large area to capture images that can be merged into an ortho-rectified photo mosaic, or orthomosaic. One of the main problems for that is to get the orthomosaic accurately georeferenced, which could be critical for decision making or to implement precise preventive/corrective actions in the crops. An accurate georeferenced orthomosaic can be obtained either by direct or indirect methods. Whereas indirect methods have been extensively studied before, there are few results reported in the literature for direct methods. In this work, the accuracy of an orthomosaic direct georeferenced using Real-Time Kinematics (RTK) applied during the UAV flight was evaluated. The aim was to obtain accurately referenced orthomosaics that could be used in the future by other autonomous vehicles. Results shown a reduction in the positioning error of up to 98% when compared with direct GPS georeferencing.