Quick and early disease detection by drones with hyperspectral camera

Idaho State University researchers have found high-tech cameras mounted on drones flying 60 meters above potato fields can effectively detect individual diseased and stressed plants.

Furthermore, greenhouse testing by University of Idaho plant pathologist Louise-Marie Dandurand, using the same advanced sensors to assess diseased plants, has confirmed specific crop ailments, such as nematode feeding or potato virus Y infection, can be differentiated by studying a plant’s “spectral signature.”

The researchers, led by ISU assistant geosciences professor Donna Delparte, intend to use those signatures to develop algorithms for quickly identifying crop ailments detected by multi-spectral and more sensitive hyper-spectral cameras.

Donna Delparte:

“We are teasing out new algorithms to automate detection so we can quickly pick out sick plants in the field.”

“Initial results are showing great promise to build out the algorithms and potentially could lead to commercialization opportunities and getting it to growers.”

Delparte’s project started during the 2014 season, with $150,000 in USDA funding. The Idaho Global Entrepreneurial Mission — an Idaho Department of Commerce initiative to commercialize technology developed through partnerships between private industry and public universities — awarded another $179,000 toward the multi-year project. J.R. Simplot Co. has assisted Delparte with soil analysis, crop expertise and grower recruitment. In 2015, the researchers used drones to gather images of about a half dozen potato and sugar beet fields from American Falls to Idaho Falls.

Allan Fetters, Simplot Technology Director:

“There’s so much new technology coming into the marketplace.”

“This is a good opportunity to work with ISU and understand how we might use this technology to bring value to farmers of Idaho.”