Evaluation of Different Flight Courses with UAV in Vineyard

Abstract

Precision farming has consolidated the use of Unmanned Aerial Vehicles (UAVs), which have confirmed their role as excellent tools for data acquisition and analysis. These systems have wide operational flexibility, good flight autonomy and excellent spatial resolution of multispectral images. They proved to be more competitive in data acquisition than satellite systems, especially on small areas typical of Italian viticulture. Unfortunately, there is little knowledge in literature about the flight course influence in the extraction of canopy spectral and biometric information. The aim of this work is to compare the different flight directions that can be performed with UAVs, to assess the differences in the acquisition of the spectral response of vineyard canopy and biometric characteristics. The vineyard is located in Camporeale, (Sicily, Italy); with a typical cultivar (Catarratto). The surface area of the experimentation field is 3.2 ha. Row orientation is NE-SW; the field can be divided into two different orographic area. In specific, one of them is flat and the other is hillside with medium-hight slope. Remote images were acquired with a multirotor UAV, equipped with a camera with six multispectral bands, the RGB for visible light images and five monochrome narrow bands for multispectral images. The sensor has a field of view (FOV) of 62.7° and 90° downwards. Front and side over-lapping were 70%. Two flight courses were carried out, one perpendicular and the other parallel to the orientation of the rows. The results show differences in the reconstructions image that induced a variability in multispectral data.