Graduate Research Assistant University of Minnesota
Disclosure: Disclosure information not submitted.
Measuring phenotypic traits is an important step in corn breeding programs and for making management recommendations. However, due to the time and labor needed for conventional data collection, phenotypic traits such as plant height are often only collected once a year, after peak growth. The ability to collect important phenotypic data throughout the growth season would be highly beneficial both for breeding and management purposes. Previous work has focused on the use of Unmanned Aircraft Systems (UAS) for collecting traits on homogenous landscapes, which are typical of breeding nursery and trial fields. We are working to optimize these previously developed methods on land more representative of production fields which consist of inconsistent soil types, water retention abilities, and elevations. To this end, we imaged a production field planted with Mycogen F2F569, a BMR corn hybrid, every two weeks using a DJI Phantom 4 Advanced Drone throughout the growing season. Point clouds were generated using Agisoft Metashape Pro and aligned using Cloud Compare. Digital surface models were then generated to obtain plant heights throughout the field. Twenty regions of the field (~6 feet x ~7 feet) that represented the variation in landscape of the field were defined as plots. Plant height from eight plants per plot was measured by hand following each flight. Accuracy of the extracted plant height compared to the manual hand measurements was analyzed on a per day basis and as growth rates throughout the growing season. Monitoring crop growth throughout the season provides an additional piece of information that can be used to assess crop performance and could be used to improve mid-season management practices by producers.