EVALUATION OF RGB-BASED VEGETATION INDICES FROM UAV IMAGERY TO ESTIMATE FORAGE YIELD IN GRASSLAND

摘要: Abstract. Monitoring forage yield throughout the growing season is of key importance to support management decisions on grasslands/pastures. Especially intensely managed grasslands, where nitrogen fertilizer and/or manure are applied regularly, precision agriculture applications beneficial sustainable, site-specific treatment, grazing and forecasting mitigate potential negative impacts. To these decisions, timely accurate information needed plant parameters (e.g. yield) with a high spatial temporal resolution. However, in highly heterogeneous communities such as assessing their in-field variability non-destructively determine e.g. adequate application still remains challenging. biomass/yield estimation, an important parameter grassland quality quantity, rather laborious. Forage (dry or fresh matter) mostly measured manually rising plate meters (RPM) ultrasonic sensors (handheld mounted vehicles). Thus cannot be assessed for entire field only disturbances. Using unmanned aerial vehicles (UAV) equipped consumer grade RGB cameras can by computing RGB-based vegetation indices. In this contribution we want test evaluate robustness indices estimate dry matter recently established experimental site Germany. Furthermore, VIs compared computed from Yara N-Sensor. The results show good correlation NGRDI R2 values 0.62.