Precision Agriculture is a hot topic these days as researchers look for ways to reduce the massive amounts of chemicals that are currently sprayed on food crops. A new research paper describes one of the ways robots may help in Early Season Site-Specific Weed Management (ESSWM). In the study, a UAV equipped with a multispectral camera collected images of sunflower field infested with naturally occurring weeds. Similar imaging techniques using traditional satellite and aerial methods to adjust herbicide distribution have yielded up to 50% reduction in the total amounts of herbicides used. The researchers hope to replicate this process using the less expensive flying robot. From the paper:
Weeds are distributed in patches within crops and this spatial structure allows mapping infested-uninfested areas and herbicide treatments can be developed according to weed presence. The main objectives of this research were to deploy an UAV equipped with either, RBG or multispectral cameras, and to analyze the technical specifications and configuration of the UAV to generate images at different altitudes with the high spectral resolution required for the detection and location of weed seedlings in a sunflower field for further applications of ESSWM. Due to its flexibility and low flight altitude, the UAV showed ability to take ultra-high spatial resolution imagery and to operate on demand according to the flight mission planned.
In the experiment, an MD4-1000 VTOL quadcopter from Microdrones GmbH was used (pictured above). The UAV was equipped with GPS, waypoint navigation software, telemetry logging, and two cameras: an Olympus PEN E-PM1 point-and-shoot digital camera and a Tetracam Mini-MCA-6 six-band multispectral camera. The immediate research goal in this project was to figure out what sensors and image processing techniques would work, so further improvements are quite likely. Now all they need to do is find a catchy name for this technology: weedbots? agridrones? herbidroids? For all the details, read the paper "Configuration and Specifications of an Unmanned Aerial Vehicle (UAV) for Early Site Specific Weed Management" by Jorge Torres-Sánchez, Francisca López-Granados, Ana Isabel De Castro, and José Manuel Peña-Barragán. Read on for some photos showing sample imaging data from the UAV and waypoint navigation paths over the test crop.