INVESTING CHECKOFF DOLLARS
SEEING BEYOND THE NAKED EYE Advancing digital imagery in soybean management BY PETER KYVERYGA
P
recision agriculture today goes above field level. Photos captured with specialized cameras high above the crops are used to get the most accurate data for optimum performance. This imagery consists of two parts: visual, what can be seen with our eyes; and digital, the invisible data behind the image. The Iowa Soybean Association (ISA) Analytics team is working with two farmers, several commercial imagery providers and university partners to test different imagery systems to best detect plant stress, predict yields and measure
22 | JANUARY 2019 | IASOYBEANS.COM
calibration quality. “Visually assessing aerial images alone is often quite limited compared to what the invisible data may reveal,” says Brad Wirt, ISA GIS specialist. “The digital aspect of the imagery is critical for analysis.” The Analytics team studies images captured by airplanes or drones that fly over two specific fields every seven to 14 days during the growing season. Calibration is important for accurate image comparison. Calibration tarps — with reflectance values from very dark (3 percent reflectance) to very bright (56 percent reflectance)
— are placed next to the fields and included in each image as a consistent baseline. The ISA Analytics team examines each image for spatial accuracy, spectral color band integrity and the ability to seam images together into one composite picture. Inaccuracies in these aspects can easily be detected by experts. However, for farmers and agronomists, spotting the inaccuracies can be challenging. For example, if the collected imagery is used to target weed spraying, an error of 15-30 feet in spatial accuracy could result in spraying the wrong areas.