Abstract
The research developed a system that was able to acquire the spectral reflectance using the fusion of machine vision and ambient illumination sensor. The research proved that to be able to acquire NDVI (Normalized difference vegetation index) with reproducibility, a proper controlling of camera parameters should be done. The developed system was attached on an unmanned helicopter. By using this system, models for moisture content estimation were made. The moisture content of wheat ear was estimated accurately. As there was a linear relation between NDVI derived from radiation of satellite images and NDVI calculated from reflectance observed from the helicopter, it was possible to convert the former to the latter by a linear regression. Using the combination of helicopter-based and satelite-based NDVIs, the estimation error of moisture content was reduced.
