Satellites locate plant diseases
July 5, 2007
Pictures of farmland can be beautiful.
But to Forrest Nutter Jr., they mean so much more.
Nutter, professor of plant pathology, has discovered how to use satellite images to detect crop diseases.
Technically, the satellites aren’t used to find disease itself.
“Crops produce leaves,” Nutter said. “The diseases remove green leaf area.”
What the images actually measure are the effects of the disease, not so much the disease itself, Nutter said. Which disease the crop has is then identified by how fast the green leaf area is being removed. For example, Asian soybean rot spreads very quickly, so it would remove green leaf area faster than other diseases.
“We can look at images and say which field is likely to have soybean rust,” Nutter said.
The way to identify disease in the satellite image is the intensity, Nutter said.
If the image is more intense, the crop is likely to be healthier, and a less intense image points to a diseased crop.
One of the problems Nutter had was to figure out a way to process and analyze the satellite images.
“We had to work out the right analysis and equation to extract information,” Nutter said.
Another problem is that the Nutter can’t receive the satellite images as soon as they are taken. After they are taken, Nutter hopes to get them in seven to 14 days.
“It is a concern of national biosecurity to obtain images in real time,” Nutter said.
This research isn’t to be used exclusively in Iowa or the U.S. Nutter said the application would be used more to monitor other countries in Africa and South America to know if a soybean epidemic is headed to the United States.
John Basart, professor emeritus of electrical and computer engineering, has been helping Nutter with the project.
“One of my emphases is image processing,” Basart said.
Basart said his part in the project was to help extract information. He said there are problems beyond the time factor with using satellites.
“The images are very expensive,” Basart said.
The satellite images cost approximately $1,000 per image. Another drawback is that the Earth rotates and the satellite may not be able to take a photo of a certain field for days.
There are other alternatives to satellites that can be used, Basart said.
One alternative is using planes with cameras on the bottom. One advantage to using planes is researchers have control over where the plane flies, but planes have disadvantages, too.
“The planes only cover small regions,” Basart said.
Another alternative is what Basart and the Space Systems and Operations Laboratory are working on right now. SSOL is using balloons with specific instruments – called a payload – attached to it to get images and perform other tasks.
The balloons being used are exactly like the ones employed by the National Weather Service. On the ground, the balloon will be 8 to 10 feet in diameter, but can expand to as much as 30 feet at low-pressure altitudes.
This is a much cheaper way of collecting information on fields.
“One balloon flight costs about $600,” Basart said.
SSOL has designed a balloon that can be popped without damaging the payload. After the balloon is “cut,” a parachute shoots out and the payload floats to the ground.
The payload is tracked by GPS and is picked up by a recovery vehicle, Basart said.
In the future, SSOL will be working on using five balloons to collect spore counts at five different altitudes.
“We want to see how many spores there are and how far they drift,” Basart said.
Knowing how far spores drift will allow Nutter and Basart to know where the crop disease may be spreading.
SSOL and its research is mostly student driven. Students enjoy this work and it is advancing the state of engineering, Basart said.
In the future, Nutter is working on combining the satellite images with balloons that monitor spore density.
“SSOL gets a big charge out of this balloon research,” Basart said.