Image based yield estimation in cotton using UAS

Aerial imagery is becoming increasingly relevant as a tool for crop data collection and crop management. As analytic power and technology continue to develop and be better understood by producers and researchers, aerial imaging will become a commonly-used production and research tool. Currently, the cost of obtaining meaningful experimental data can be prohibitive, considering costs of equipment, technology, software, and labor. Furthermore, in regions where resources are scarce, high-end technology is completely unobtainable. This has motivated an effort to document and test a combination of drone hardware and software technology that is cheap, effective, and capable of generating meaningful outputs for producers and researchers. Open-source imaging software coupled with off-the-shelf consumer-grade drone hardware is neither cost-prohibitive nor difficult to obtain. This research aims to evaluate cheap, consumer-grade hardware and open-source software for its usefulness in studying crop growth and phenology in a rain-grown production setting. Specifically, the parallel objectives are assessing this technology’s potential for taking in-field seeded cotton yield measurements, and developing best-practices and an open-source framework for collecting, analyzing, distributing, and storing aerial image data.