A Workflow for Extracting Plot-level Biophysical Indicators From Aerially Acquired Multispectral Imagery

The use of unmanned aerial vehicles (UAVs) for precision agriculture and research has been increasing in recent years. Advances in technologies associated with UAVs have allowed researchers, farmers, and agribusinesses to incorporate UAVs coupled with various imaging systems into data-collection activities and aid in the decision-making process. Multispectral imagery allows for a quantitative assessment of biophysical indicators or plant health status. High spatial and temporal resolutions allow for a vast amount of data to be gathered on-demand. Data generated frommultispectral, aerialacquired imagery can complement ground-based measures of plant health status or crop performance (i.e., disease severity, incidence, vigor, stand count, stress, and yield). Remote-sensing techniques for landscape-level imagery have been well-established (Hatfield and Pinter 1993; Jackson 1984; Rock et al. 1986; Rouse et al. 1974). What is needed now is a standardized workflow for the quantitative assessment of plant biophysical indicators using low-altitude, high-resolution, UAV-acquired, multispectral imagery. With this need in mind, the authors developed and proposed a workflow (Fig. 1) for extracting plot-level vegetationindexmeans utilizing orthomosaics generated from amultirotor UAV coupled with a scientific-grade multispectral camera.