High-throughput phenotyping of wheat ear surface area and ear density in the field

Ear density (D_e) and ear surface area (S_e) in cereals are important traits for adaptation to low inputs and climate change. Here we propose a high-throughput field phenotyping method to estimate these traits using nadir and 45° RGB images acquired by the Phenomobile ground robot. First, the YOLOv5 ear detection algorithm is applied to nadir RGB images to estimate D_e. Second, an ear segmentation algorithm is applied to nadir and 45° RGB images to compute the ear gap fraction at different viewing angles. The Beer-Lambert law is then inverted to compute the ear area index (EAI) from the observed ear gap fraction. S_e is finally derived as the ratio between EAI and D_e.

We applied the methodology to a panel of 10 commercial bread wheat varieties to analyse how both traits vary across 12 environments. The relative error obtained for awnless varieties is 12% (56 ears m⁻²) for D_e and 18% (1.3 cm²) for S_e. For awned varieties, ground-truth observations of S_e were shown to be biased due to an overestimation of awns contribution, leading to an error of 41% (3.6 cm²). S_e was strongly correlated with grain dry mass per ear at harvest (r² = 0.80 across genotypes and environments, r² per genotype ranged between 0.80 and 0.95) and EAI was strongly correlated with grain yield (r² = 0.83). These results indicate that both EAI and S_e can be interesting non-destructive proxies for yield and grain dry mass per ear.