Analysis of soybean varieties maturation using multispectral data

Abstract

Aerial photography using unmanned aerial vehicles and a multispectral camera is used to monitor the condition of soybean crops, predict their development and phenotyping. This research focused on evaluation of soybean varieties in the initial ripening period using multispectral data from an unmanned aerial vehicle. The quadrocopter Matrice 200 v2 and the MicaSense Altum multispectral camera were used to collect multispectral data. The article analyzes ten soybean varieties according to five vegetation indices (NDVI, NDRE, ClGreen, GNDVI, MCARI) and five spectral channels with wavelengths from 475 nm to 842 nm. Using the Student’s t-test, significant differences between soybean varieties and the standard soybean variety for the Orel region were revealed. Recommendations have been formed for the assessment of soybean varieties by the maturation rate using multispectral aerial photography: it is recommended to use the vegetation index GNDVI, as well as a spectral channel with a wavelength of 475 nm.