Can Peanut Agronomic Characteristics be Estimated from an Uav Platform?.

Field evaluations using low-level aerial imaging platforms are expected to relieve the phenotyping bottleneck by allowing more accurate selection of phenotypes with desirable traits and in less time than traditional methods. Recently, we initiated a study for the evaluation of suitability of RGB and hyperspectral imaging taken from an unmanned aerial vehicle (UAV) platform for peanut variety differentiation and estimation of agronomic characteristics. RGB-derived green vegetation indices (proportion of pixels with Hue values from 60 to 120°) were significantly (R²=0.40, n=28) correlated with yield of 28 peanut cultivars and breeding lines grown under water deficit in field tests under rainout shelters in Suffolk, VA. The genotypes with the highest green indices also produced the highest yields under drought. Canopy temperature and Hue angle color space characteristic derived from aerial infra-red and RGB images were also significantly correlated (R²=0.52; R²=0.73, respectively) with defoliation (measured by visual rating) due to leaf spot disease. Peanut genotypes resistant to leafspot disease, and therefore with less defoliation) were cooler than those susceptible to the disease; the more resistant genotypes also had canopies with higher Hue angle (greener canopies).