Project Details:

Small Unmanned Aircraft thermal infrared imaging system to identify soybean drought tolerant varieties

Parent Project: This is the first year of this project.
Checkoff Organization:Kansas Soybean Commission
Categories:Breeding & genetics, Water quality & management, Analytical standards & measurements
Organization Project Code:1779
Project Year:2017
Lead Principal Investigator:Ajay Sharda (Auburn University)
Co-Principal Investigators:
Daniel Flippo (Kansas State University)
William Schapaugh (Kansas State University)
Deon van der Merwe (Kansas State University)

Contributing Organizations

Funding Institutions

Information and Results

Click a section heading to display its contents.

Project Summary

Agriculture annually consumes approximately 80% of ground and surface water in the United States (U.S.) (USDA, 2013). Recent decline in water availability and droughts are becoming critical factors impacting crop yield goals in the U.S. This changing crop production scenario requires drought tolerant crops to sustain crop production goals for national and global food security. Therefore, there is a need for sensing systems to accurately and rapidly identify lines with higher degree of drought tolerance to develop varieties with greater resilience to drought exposures. Thermal sensing approaches have been investigated because they are non-contact and less labor intensive, and offer non-destructive monitoring to assess crop stress from leaf canopy temperatures (Grant, et al., 2006). Crop canopy temperature has been accepted as an indicator of crop water stress (Ballester, et al., 2013; Grant, et al., 2006; Jones, 1999). Lightweight Thermal Infrared (TIR) cameras can spatially map temperatures via a thermal image to measure subtle, heterogeneous characteristics of leaf dynamics (Liu, et al., 2011). Each image is comprised of pixel arrays representative of pixel resolution of the TIR camera. TIR imagery and derived indices could allow researchers to conduct individual crop temperature profiling (Leinonen & Jones, 2004); sense crop health status (Taghvaeian, et al., 2013) and develop correlation among crop canopy temperature, stomatal conductance, and leaf water potential (Hackl, et al., 2012). The TIRIS can capture canopy temperature profiles at a high temporal and spatially resolution (Mangus et al., 2015). sUAS equipped with TIRIS can be flown at low altitude to develop spatial canopy temperature maps or thermograms to assess enhancements in crop development, tolerance to drought among different genotypes and yield improvements. This approach will provide a high throughput phenotyping platform and enhance capabilities to select not only drought but in future disease and pest resistant varieties as well.

Project Objectives

1. Record in-season soybean canopy thermal profiles using thermal infrared imaging system (TIRIS)

2. Evaluate the effectiveness of whole-plot thermal imaging for high-throughput phenotyping

Project Deliverables

To achieve the overall goal, the proposed work consists of four major steps: 1) collecting aerial imagery using Small unmanned aircraft system (sUAS) platform on target plots for desired spatial resolution, image overlap, and altitude, 2) image analyzing using thermographic calibration curves compensating for environmental conditions during sensing campaigns, 3) image stitching (mosaicking) to create whole-plot spatial canopy temperature maps, and 4) measuring agronomic traits on a diverse range of soybean genotypes (including plant introductions and breeding lines). sUAS TIRIS has been developed at Biological and Agricultural Engineering department. It utilizes uncooled thermal camera and other required sensors for environmental compensation. Deployable temperature reference panels will be developed for in-flight accuracy assessment. These reference panels will be used to initialize and calibrate camera before, during, and after each flight. Multiple temperature reference panels will use thermistors connected to a data acquisition system (DAQ) to record real-time reference panel surface temperature while monitoring environmental parameters (i.e., air temperature, humidity, solar radiance, wind speed, and direction). As tested in full-season greenhouse corn studies last year, the TIRIS system will utilize calibration techniques using reference calibration images to quantify canopy temperatures from camera pixel intensity; corresponding transfer function from pixel intensity to surface temperatures along with camera sensor settings to accurately measure crop temperature profiles. In addition to thermal imaging, color infrared camera (CIR) mounted on the sUAS will be used to capture canopy reflectance. Ground reflectance reference panel will be used to calibrate CIR camera at the flying altitude.

TIRIS will be integrated with multi-rotor sUAS to collect aerial imagery. Test will be conducted at Kansas State University’s soybean breeding trials. A Certificate of Authorization (COA) for one year of flying sUAS for these locations will be obtained through K-State Saline Unmanned Aircraft Systems Program Office (USAPO). The research plots will consist of irrigated and dryland plots near Manhattan and Salina, Kansas. The target genotypes to be monitored will include lines that are under evaluation and development in the abiotic stress soybean breeding and genetics program pipeline. The sUAS with TIRIS will be flown at 15 and 25 m. Different heights are selected to understand and differentiate the effect of flying altitude on thermal imagery accuracy and image pixel degradation, dynamic accuracy of imaging system at different altitudes, pixel and canopy temperature resolution, assess productivity in terms acres covered per hour and assess economics of using this system. Nine airborne campaigns at each site will be conducted over the experimental plots to capture images around solar noon flight times to investigate the physiological status of the plant capturing the key flowering and seed-fill phases. In addition to aerial imagery, sampling regions will be randomly selected for ground truthing and; plant health and growth related data collected. Actual canopy temperature will be measured with leaf clip radiometers to validate aerial imagery results at selected sampling locations. Ground based evaluation of agronomic and physiological traits such as chlorophyll content using SPAD will be conducted for analysis and screening of drought resistant varieties. Plant material measurements will include seed yield, maturity, lodging, height and visual wilting scores. Potential effectiveness of this system to screen genotypes will be assessed by correlating thermal canopy profiles with relative seed yield, maturity and response to drought and heat stress.

Progress of Work

The plots have been established at Manhattan and Salina for summer 2017 data collection. The data collection will be conducted from July 15th through September 15th. The graduate student has accepted the position and is expected to join from Fall 2017.

Updated June 21, 2018:

View uploaded report PDF file

Updated June 21, 2018:
The plots have been established at Manhattan, Rossville, Wamego, and Salina for summer 2018 data collection. The data collection will be conducted from July 15th through September 15th.

Updated July 2, 2018:

View uploaded report PDF file

Final Project Results

Benefit to Soybean Farmers

Rapid identification of drought resistant varieties could enhance the capabilities of producers to more efficiently used available water resources and grow better crops even in deficient irrigation/moisture conditions

Performance Metrics

Project Years