Eye On Cotton Stands

Plant populations can be monitored from above.

Image by Karl Wolfshohl

Stand establishment can be a challenge in cotton, and evaluating plant populations is the first step to taking proactive measures. Research is showing putting an unmanned aerial system (UAS), often called a drone, in the air could be an option to the spring chore.

Walking fields to sample stands is reliant upon a highly uniform plant population across the entire field and can be influenced by human bias, says Shawn Butler, graduate student in the University of Tennessee (UT) College of Agricultural Sciences and Natural Resources. “Theoretically, an aerial approach could provide spatially dense information on plant populations across large areas quickly and remove human bias,” he adds.

For the past two years, Tennessee researchers assessed plant stands of emerging cotton through manual counting and through images obtained from both digital and multispectral cameras mounted beneath a quadcopter. The quadcopter was flown at varying altitudes ranging from 30 to 120 meters.

P[] D[x] M[x] OOP[F] ADUNIT[] T[]

Tyson Raper, project leader and assistant professor with the UT department of plant sciences, says the jury is still out on the perfect observation height. The lower altitude makes it easier to detect plants but greatly reduces the number of acres that can be covered.

EYE IN THE SKY. Of the two camera systems analyzed, the images produced from the multispectral camera proved to be more accurate in estimating plant populations, with a greater-than-93% accuracy. However, researchers say the red, green, blue (RGB) images produced by the less-expensive digital camera still looked promising, with a greater-than-85% accuracy using current methods and scripted programming.

“Based on initial results, the aerial imagery provided by either RGB or multispectral sensors may be a sufficient tool to improve accuracy and efficiency of plant stand assessment,” Butler says.

Raper says that the multispec cameras are more expensive but measure reflectance in the red, blue, green, red-edge and near-infrared wavelengths. “Healthy, photosynthetically active plants reflect energy much differently in the near-infrared and red-edge regions. Quicker processing and greater accuracy with the multispec camera may justify the additional cost, especially when divided over large acreages.

“Crop monitoring is a big obstacle for many producers,” he adds. “We want to continue to evaluate tools and methodologies that have the potential to help farmers overcome monitoring challenges, improve response time and increase profitability.”

The study was conducted at three locations: the West Tennessee AgResearch and Education Center, in Jackson; and the UT AgResearch and Education centers, at Milan and Ames Plantation. The other coproject team leader included Mike Buschermohle, interim assistant dean of UT Extension. Cotton Inc. provided partial support for the project.

[PF_0118]

Past Issues

and