Bison‐Fly: An open‐source UAV pipeline for plant breeding data collection
Filipe Inácio Matias, Andrew Green, Jennifer Lachowiec, David LeBauer, Max Feldman
Abstract
Abstract Bison‐Fly is the open‐source pipeline of unoccupied aerial vehicle (UAV) applications for plant breeding developed by the Spring Wheat Breeding Program at North Dakota State University in partnership with the Drone2Phenome community. We present a step‐by‐step pipeline to collect, process, and apply UAV data for plant breeding to facilitate data processing, reduce errors, and create new traits. This open‐source R code can be easily adapted for different crops and improved according to user requirements. In addition, Bison‐Fly provides a free and complete dataset with (a) RGB, Multispectral, and Digital Surface Model orthomosaics from 15 time points throughout the season; (b) field trait data with experimental design; and (c) daily metadata with soil and weather information. We hope this pipeline helps guide students, researchers, and breeders on day‐to‐day activities. Bison‐Fly is available at https://github.com/filipematias23/Bison‐Fly .