Farm Progress is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Serving: United States

''Seeing'' weeds

How would you like to spray 25%, 50%, maybe up to 75% less herbicide on your soybeans using spot spray application based on weed detection sensors, and still get the same yield that a broadcast application would bring?

That's the percentage of herbicide savings John Hummel, ag engineer with the USDA-ARS, Columbia, MO, has been achieving using Patchen's WeedSeeker in tests on soybeans.

Other researchers are achieving similar herbicide savings using prototype smart sprayers that detect weeds.

The WeedSeeker uses an optical sensor to measure infrared and near-infrared reflectance and the presence of chlorophyll. It can detect any green plant, but it cannot distinguish between crops and weeds. When the WeedSeeker detects a green plant, it activates a solenoid that turns the sprayer on and off. For postemergence applications in row crops, the WeedSeeker sprayer is equipped with hoods between the rows. When the sensor detects a green plant between the crop rows, the sprayer is activated and the hood keeps the herbicide away from the growing crop. The WeedSeeker sprayer can be equipped with two tanks so a selective herbicide can be applied over the row at the same time that a nonselective herbicide is applied between the rows.

In row crops the WeedSeeker is being used primarily in cotton. It's also being used in orchards, vineyards and roadside applications.

Significant savings

In his tests using the WeedSeeker in corn and soybeans in Illinois, Hummel found that yields were statistically the same whether they used the sensor or broadcast the herbicide under the hood. “In the years where we were able to get into the field in a timely basis we saved some herbicide using the sensor,” he says. “In soybeans we had herbicide savings ranging from 20 to 75%. In corn, 20 to 25% savings were typical.” Hummel used a tractor guidance system from Tri-R-Innovations to stay between the rows at all times.

James Hanks, another ag engineer with the USDA-ARS in Stoneville, MS, has tested the WeedSeeker in cotton, soybean and corn crops for several years to evaluate hooded sprayer applications with and without sensors. “The average postemergence herbicide savings for cotton was 73% over three years,” Hanks says. “In soybeans the herbicide savings averaged 49%. During a one-year study on about 2,500 acres of a commercial farm in the Mississippi Delta, the savings were 71% in cotton, 53% in corn and 70% in soybeans. Weed control equaled that of spraying the herbicide continuously under the hood.”

Oklahoma State University has a prototype smart machine that uses the WeedSeeker sensor technology to read wheat plants' nitrogen needs and spray fertilizer at a variable rate on the go. Preliminary results indicate that nitrogen use efficiency can be increased from 50 to 70%. The machine was also used to spot spray herbicide only where there were weeds during fallow conditions. Herbicide use was reduced about 50%, according to John Solie, an ag engineer at the university.

An 8-row, 30-in.-row WeedSeeker unit with a fixed boom costs about $23,000. For more information, contact Patchen Inc., Dept. FIN, 740 S. State St., Ukiah, CA 95482, 888/728-2436,

Discriminating sprayer

Several manufacturers are looking at a prototype sprayer developed at the University of Illinois. The “smart sprayer” uses video cameras mounted ahead of the spray boom to determine the size and density of weeds between the crop rows and creates a weed map and application prescription. The two-dimensional image sensor array has higher resolution than the WeedSeeker's photo sensor.

The University of Illinois smart machine sprays a 10% rate of herbicide across the entire field but automatically calculates and increases the rate at each spray nozzle to match weed size and density. “Young weeds may only require a 20% rate of the herbicide, whereas areas with large numbers of large weeds may need a full rate,” explains Dr. Lei Tian, professor of ag engineering at the University of Illinois. The machine can apply a different variable rate of herbicide from each nozzle, whereas the WeedSeeker nozzles can only be turned on or off.

“We've found that many fields are only 10 to 30% infested with weeds that need to be sprayed. Our average herbicide savings has been 52%,” Tian says. He has been using Roundup Ready herbicide on Roundup Ready crops, but a selective herbicide could be used on conventional crops instead. Like the WeedSeeker, the University of Illinois machine does not distinguish between crops and weeds or different weed species at this time.

For more information, contact Dr. Lei Tian, University of Illinois, 1304 W. Pennsylvania Ave., Urbana, IL 61801, 217/333-7534, e-mail

More research

Researchers at the University of California — Davis have a prototype machine vision sprayer that takes color video pictures to develop a custom spray map. The prototype can distinguish between broadleaves and grasses. Individual plant leaves can be sprayed at the centimeter level, according to David Slaughter, ag engineer. It is currently being tested in cotton.

Researchers also are working on algorithms that will help sensors detect and identify weeds using near-remote and remote sensing devices. Some are using spectral signatures, whereas others add textural and shape analysis to help distinguish between plants. Case Medlin, an agronomist with Purdue University, believes researchers will be able to identify spectral reflectance signatures unique to specific weeds and crops. He and other researchers are trying to overcome the challenges of using sensors in a field setting where crop leaves and weed leaves overlap, wind blows leaves on their underside, and lighting and moisture conditions change.

Another goal is to use real-time sensors for multiple purposes and thus bring down the cost. As in the Oklahoma example, the sensor could measure chlorophyll content of corn leaves to determine rates of side-dressed nitrogen. In addition, the sensor might count emerged corn plants or recognize insect infestations.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.