Determining Crop Growth Dynamics in Sorghum Breeding Trials Through Remote and Proximal Sensing Technologies

With the advent of sophisticated spectral imaging the development of proximal and remote-sensing high-throughput field phenotyping platforms has become progressively feasible. Such platforms, combined with new innovative image-analysis and machine-learning tools, will likely become the benchmark in high-throughput plant phenotyping frameworks for determining plant responses at the leaf, plant and canopy level. At the University of Queensland, we have developed a cost-effective high-throughput phenotyping pipeline that harvests sensing data from plants utilising proximal sensors. Here, we describe the application of this software pipeline to analyse crop growth dynamics from (i) proximal sensors on-board a tractor-based phenotyping platform and (ii) cameras attached to small unmanned aerial vehicles. Specifically, we discuss the use of high-resolution characterization of time-sequence red-edge data obtained from both sensing platforms to derive estimates for dynamic growth parameters in sorghum breeding trials. Although some temporal bias existed between the two platforms the relationships at spatial peak canopy (R
²
= 0.71) and plot level (R
²
= 0.89) were strong. Application of these technologies across breeding plots will enhance phenotyping capabilities and hence the ability to discriminate among responses of genotypes across different environments.