Corn and soybeans, two essential crops that can be processed into a myriad of foodstuffs, and are eaten the world over. Like any crop, corn and soybeans are subjected to stress from any number of reasons, including drought, flooding, poor soil conditions, even the absence of natural pollinators. Being able to identify plant problems early can mean the difference in saving the crop or losing it. Still, a lot of farms don’t have access to the necessary hardware to monitor acres of fields, as most solutions can be costly.
To overcome that problem, researchers from NC State has developed an inexpensive camera system that costs less than a smartwatch and can monitor crop stress remotely. The system can help researchers examine ways to make agricultural systems more resilient, plant breeders that can produce more drought-tolerant plants and could alert farmers when to irrigate their crops.
The team's StressCam is simplistic in design and features a Raspberry Pi and camera that takes pictures of fields, which is set on a timer that turns the system on in the morning and off during the evening. For cornfields, the camera is mounted at a 90-degree and takes images every 30 minutes to watch for curling leaves, a sign of stress in corn crops. For soybeans, the camera is positioned at a 45-degree angle and takes photos every 15 minutes to look for wilting. The system runs on solar power, and backup batteries are used on cloudy days.
The Raspberry Pi is programmed with a machine learning algorithm that analyzes those images for signs of stress, which then uploads to IBM’s cloud platform for farmers, researchers, and plant breeders. The researchers have built 50 StressCams and plan on deploying them over corn fields in Maryland, at the Beltsville Agricultural Research Center, and other research stations for testing, with the ultimate goal of deploying them to farms all across the US.