UAV-based sunflower yield prediction using multispectral vegetation indices

Sunflower yield varies spatially and temporally depending on factors such as weather, altitude, seed variety, plant density, available water, nutrients, and sowing date. These are the primary factors influencing crop yield. With the use of unmanned aerial vehicles (UAVs), temporal resolution can be adjusted according to the user's needs, while spatial resolution depends on the capabilities of the sensor and flight altitude. In this study, vegetation indices derived from multispectral camera imagery were used for yield estimation through a linear regression model. Vegetation indices such as NDVI (Normalized Difference Vegetation Index), MCARI (Modified Chlorophyll Absorption in Reflectance Index), SAVI (Soil-Adjusted Vegetation Index), CIRE (Chlorophyll Index Red Edge), LCI (Leaf Chlorophyll Index), and GNDVI (Green Normalized Difference Vegetation Index) were calculated using green, red, red-edge, and near-infrared (NIR) bands. The values of these indices were obtained from UAV flights conducted on five different dates in the study area. In the regression model developed using the NDVI4 index value for the R-5 growth stage of the plant, yield estimates were obtained as 336.72 kg for Test Area 1, 381.77 kg for Test Area 2, and 400.62 kg for Test Area 3.