UAV imagery for bread and durum wheat intro-selection lines selected for root traits

Full description

A total of 44 bread wheat intro-selection lines (BC2F4:6) in four genetic backgrounds (Borlaug100, Kingbird, Mace and Suntop) and 11 durum wheat intro-selection lines in the DBA Aurora background were phenotyped for canopy development in 2021 at Gatton Research Farm, Queensland, Australia (27°32'34"S; 152°19'59"E). Lines were evaluated in full-sized plots (1.52m wide x 6m long) and the design for this experiment followed a row-column split-plot design with destructive and non-destructive treatments to enable destructive root coring at flowering (4 reps destructive and 4 reps non-destructive). Weekly UAV flights were conducted using a Matrice 300 RTK-DJI UAV fitted with a ‘MicaSense Altum-PT’ multispectral and thermal high-resolution camera. UAV flight altitudes were set at 20m and carried out on clear and still days between 10:00 to 11:00 AM local time to obtain high resolution raw images which were processed and stitched using Agisoft Metashape software. A single geo-referenced orthomosaic TIF image was generated per flight consisting of six channels (Blue, Green, Red, RedEdge, Near Infared and thermal). Shapefiles containing single plot polygons were created in ArcMap to extract plot-wise values in a matrix. Imagery from 24 flights are provided, including the raw images from both cameras as well as the Metashape project file which enables future editing of the project using Agisoft Metashape software. The output folder contains the digital elevation model (DEM) for all flights, orthomosaic folder for all flights (Orthos) and the reports for checking the quality of the flights and stitching (Reports) and excel file (Gilbert_03_GCPs) for the GCPs used for optimising image alignment during the stitching process.