A research team at the University of Nebraska-Lincoln Panhandle Research and Extension Center is developing a method of scheduling irrigation for dry edible beans that uses a sensor to remotely check the temperature of the crop canopy to determine when the crop needs water.
The method, known as infrared radiometry thermometer (IRT), has the potential to be cheaper and easier to install and manage than soil-moisture probes, according to Xin Qiao, Irrigation and Water Management Specialist at the Panhandle Center.
Qiao and his team are testing sensors and developing software that will be key to the IRT system. One part of the software will analyze digital RGB images from the field to estimate the percentage canopy cover. (RGB stands for Red Green Blue, a color mode used in digital images.)
Research so far has shown that IRT works for dry beans. Qiao said this method has been used for years in corn fields, but this project would adapt it for use in dry edible beans.
When used on corn fields, the IRT sensors are pointed 45 degrees across the rows, and see only foliage when corn canopy are 80 percent covered. Since beans are shorter than corn, and planted later in the season, an infrared thermometer sensor is pointed at a downward angle and parallel with planting rows, and it sees both bean foliage and soil.
It is essential to develop software and algorithm to differentiate between foliage and soil since soil is “noise” in the measurements. Some of the software being developed will recognize the difference between bean foliage and soil. That task is being undertaken by Weizhen Liang, post-doctoral research associate at the Panhandle Center, and Kelly McCoy, an intern who is a senior at Scottsbluff High School. The project is funded by the Nebraska Dry Bean Commission.
According to Liang, the software estimates the percentage of crop canopy cover, a key parameter to estimating the dry bean evapotranspiration (crop water use) for irrigation scheduling.
The software uses digital images (RGB) from a cell phone or camera to quantify the percentage of canopy cover. This parameter will then contribute to the calculations that estimate canopy temperature and soil temperature, according to Liang.
Using IRT to determine whether to irrigate dry beans can be simpler than using soil moisture sensors, because soil probes need to be installed just right and calibrated to the specific soil type in order to work correctly, according to Qiao. Each one must fit very snugly in its hole in the ground. And probes measure soil moisture in a 1- to 2-inch area around each probe, compared to a 23-foot square area for an IRT sensor under current set up.
Qiao said their use of IRT sensors in corn has been studied since 1981. But nobody has calculated full-season crop water use for beans using IRT system.
Liang has been working on this method since 2015, when she was at Clemson University, using a similar method on peanut and soybeans. They will also adapt the same concept to sugarbeets, another crop that is low to the ground compared to corn, under a funded project by Western Sugar Cooperative.
According to Liang, to write the software it was necessary separate the color groups in field photos to identify colors of foliage, soil, and irrelevant “noise.” She showed McCoy how to use statistical analysis to separate all this group, and the intern performed the time-consuming work, which involved calculating the red, green and blue digital values that composed each of the colors.
The software will produce one parameter, the percentage of canopy coverage compared to soil. This will help the system measure the crop canopy temperature, which is related to the plants’ stress and water requirement. This will tell the irrigator whether the plant needs to be irrigated.
McCoy, a senior at Scottsbluff High School who hopes to study electrical engineering in college, said SHS has an internship program, but none of the internships in the program were a good fit for him. So he talked to his parents and sought out other opportunities, which led him to the Panhandle Center and Qiao’s irrigation scheduling project for dry edible beans.
Working on the IRT irrigation sensor project “kind of opened my eyes to the fact that there’s a whole ag side to software development,” he said. One of his older brothers studied a similar field, and each of his three brothers worked at the Panhandle Center, but Kelly said he wasn’t aware of how many jobs are related to agricultural software. “I didn’t realize how many ag-based occupations in this field were as interesting as they are.”
The current project will last until the summer.
McCoy said he learned some computer code languages in school such as Python and HTML, but the irrigation sensor project required him to learn a language new to him, Visual Basic.
According to Qiao, the goal of the project this year is to develop a website to use with the IRT sensor system. If a farmer shoots a photo of a bean field, the website will fetch data from the nearest weather station and combine it in a crop-water model to provide crop water use.