Yield and leaf area index estimations for sunflower plants using unmanned aerial vehicle images
Özet
Vegetation is commonly monitored to improve efficiency of various agricultural practices. Spatial and temporal changes in plant growth and development can be monitored with the aid of remote sensing techniques employing ground, aerial, and satellite platforms. Unmanned aerial vehicles (UAV) and multi-spectral cameras developed for UAVs have an important potential for agricultural management activities with high-resolution spatial and temporal images. However, UAV images should be assessed based on ground measurements for using these images as a decision-support tool in agriculture. This study was conducted to estimate sunflower leaf area index (LAI) and yield with the aid of Normalized Difference Vegetation Index (NDVI) images generated from raw UAV images. Furthermore, UAV-based NDVI values were compared with NDVI values calculated by using hyper-spectral measurements carried out with a ground-based spectroradiometer. Between July and August of 2017, six flight missions were conducted and spectral measurements were made simultaneously. A significant correlation (R-2=0.77) was determined between NDVI values that belong to UAV platform and spectroradiometer. Also, regression models developed for sunflower LAI and yield estimation depending UAV-based NDVI have R-2 values of 0.88 and 0.91, respectively.
