Comparing RGB - based vegetation indices from UAV imageries to estimate hops canopy area

Remote estimation of hops plants in hop gardens is imperative in field of precision agriculture, because of precise imaging of hop garden structure. Monitoring of hop plant volume and area can help to predict the condition and yield of hops. In this study, two unmanned aerial vehicles (UAV) – eBee X senseFly UAV equipped with Red Green Blue (RGB) S.O.D.A. camera and Vertical Take-Off Landing (VTOL) UAV FireFly6 Pro by BirdsEyeView Aerobotics equipped with MicaSense RedEdge MX camera were used to acquire images of hop garden at phenology stage maturity of cones (24 th July) before harvest. Seven commonly used RGB vegetation indices (VI) were derived from these RGB and multispectral (MS) images after photogrammetric pre-processing and orthophoto mosaic extraction using Pix4Dmapper software. Vegetation Indices as the Green Percentage Index (G%), Excess of Green Index (ExGreen), Green Leaf Index (GLI), Visible Atmospherically Resistant Index (VARI), Red Green Blue Vegetation Index (RGBVI), Normalised Green Red Difference Index (NGRDI) and Triangular Greenness Index (TGI) were derived from both data sets. Binary model from each of VI was derived and threshold value for green vegetation was set. The results showed significant differences in hop plant area based on the specifications of cameras, especially wavelengths centres, and design and flight parameters of both UAV types. The comparison of various indices showed, that ExG and TGI indices has the highest congruity of estimated vegetation indices in hop garden canopy area for both used cameras. Further processing by Fuzzy Overlay tool proved high accuracy in green canopy area estimation for ExG and TGI vegetation indices