Unmanned Aerial System (UAS) for Precision Agriculture: First Results from a Growing Cycle of Cotton

Mounted with various cameras and sensors, a low-cost Unmanned Aircraft System (UAS) offers us a unique opportunity to develop smart applications for precision agriculture with high spatial, spectral and temporal resolutions. We are currently developing a UAS-based sensor platform that includes various sensors such as a high precise Global Positioning System/Inertial Navigation System (GPS/INS) system, an optical camera, a multispectral camera, and a thermal camera. The first version of the platform has been utilized to monitor a growing cycle of cotton in 2015. About 150 aerial images were captured at a weekly temporal resolution in a test field of 50x90 meters. The UAS flied at a very low altitude of 15 meters in order to achieve a high spatial resolution of 7mm/pixel. Images collected in a weekly fly mission were processed to generate 1) a geo-referenced mosaicked image that can be used to estimate the plant coverage during a growing cycle of cotton, and 2) a 3D canopy model that can be used to estimate the plant heights. The growing status of the cotton plants has been assessed on a weekly basis by evaluating two parameters: plant coverage and plant height. Four indexes have been estimated for each crop grid including: plant coverage, mean of the plant heights, variance of the plant heights, and maximum plant height. The growing status in terms of plant coverage estimated with the UAS-based system agreed very well with that determined with the leaf area index. Our experience from monitoring one growing cycle of cotton shows that UAS has a great potential for precision agriculture applications.