2019
Crop Scouting Using Aerial Imagery in Delaware Soybean Fields
Contributor/Checkoff:
Category:
Sustainable Production
Keywords:
Field management Nutrient managementSoil healthTillageYield trials
Parent Project:
This is the first year of this project.
Lead Principal Investigator:
Jarrod Miller, University of Delaware
Co-Principal Investigators:
Bill Cissel, University of Delaware
Mark VanGessel, University of Delaware
+1 More
Project Code:
Contributing Organization (Checkoff):
Institution Funded:
Brief Project Summary:

For soybean production, drones have been used to estimate yields, detect disease, and manage weeds. However, most of this research is performed with multispectral cameras and specialized software, which may not be practical for row crop farmers. With less advanced equipment, common tasks like crop scouting, observing irrigation systems, or monitoring livestock can be performed. Drones may provide trained agronomists and crop advisors scouting crops additional assistance in this field. The goal of this project was to compare traditional versus drone scouting methods to see if they could increase efficiency in either reduced time or increased issue discovery.

Key Benefactors:
farmers, agronomists, Extension agents

Information And Results
Project Deliverables

1) Estimates of time spent or saved using UAVs to scout soybeans
2) Estimates of savings if scouting solved additonal issues

Final Project Results

Updated March 18, 2020:
For this project three soybean fields (one from each county in Delaware) were selected and followed through the growing season. These fields were scouted by walking the rows as well as with aerially by drone. For walking methods, fields were covered in a zig-zag pattern at the beginning of the season, until full canopy reduced maneuverability. Following full canopy, walking could only be efficiently done along irrigation wheel tracks.

For aerial scouting, two drones were used, the Parrot Anafi (Paris, France) or the DJI Mavic Air quadcopter (Shenzhen, China), based on availability the day of the flight. Two aerial scouting methods were used in each soybean field. One was an automatic, pre-planned flight using free available software (Pix4DCapture, Switzerland) to setup a consistent scouting plan (Figure 1). The software was downloaded onto an Apple Ipad, which included celluar for global positioning (GPS) capabilities. Each field was selected by using the address and aerial photo with Pix4DCapture. The field boundary was drawn as a polygon to fit the exact outline and reduce the amount of battery used.

Height above ground level (AGL) was set for 200 feet and overlap was initially set to 50%, before being reduced to 30% after the first flight. Camera angle can be set from 0 (straight ahead) to 90° (straight down) in Pix4DCapture. For this project, camera angle was set to 45° to capture more of the field in one image, also cutting down on flight time. Images were downloaded by Wi-Fi on the first flight, and then downloaded directly from the SD card onto a laptop for subsequent flights, due to time constraints.

The second scouting method using a drone was a manual flight using the software associated with each drone. The Parrot Anafi was flown with Parrot FreeFlight6 (Paris, France) and the DJI Mavic Air was flown using the DJI GO 4 app (Shenzhen, China). Each app was downloaded onto an Ipad tablet for a larger screen. The drone was initially taken up to 400 feet to get an overall view of the field, before flying to problem areas and performing lower height scouting (Figure 2). All images were downloaded and stored on a 1TB external hard-drive.

This simple experiment was revealing when evaluating the use of drones in soybean scouting, so that outreach to farmers and consultants could be clear on the expected uses. Most of the outcomes of this project are expected; including less time spent scouting fields with a drone. The final report is attached.

View uploaded report PDF file

Drones are the latest technology to be hailed as the future of precision agriculture, but their practical use still needs to be explored. For soybean production, drones have been used to estimate yields, detect disease, and manage weeds. However, most of this research is performed with multispectral cameras and specialized software, which may not be practical for row crop farmers. Instead, with less advanced equipment, common tasks such as crop scouting, observing irrigation systems, or monitoring livestock can be performed.

This project examined the role an off the shelf, consumer variety drone could play in scouting crops. Three soybean fields were selected and followed through the 2019 growing season by either walking the field or using a drone to scout. The drone was flown by two methods, an automatic flight plan and manually guiding the drone to locations in the field.

Drone imagery was useful early in the season, however without much canopy closure, it was difficult to see field patterns. Patches of weeds were noticeable early in the season, but deer damage or seedling emergence was better observed by walking the field. Later in the season, with soybean canopy closure, field patterns of nutrient issues, deer paths, and disease pressure became more apparent. The drone was a faster tool for a scout to observe these issues and their extent. Alternatively, it was more difficult to spot weed pressure with a consumer drone camera later in the season, as they blend into the soybeans much better.

Much more time was spent walking the field than using a drone to scout, and drones were particularly helpful later in the season. With canopy closure it becomes difficult to cross the field and find issues, even when following irrigation wheel tracks.

As a scouting tool drones can be very useful to an agronomist or farmer, pending action is taken when any issues are found. The aerial view of the field gives a better vantage point to pick up the extent of any issues, especially later in the season.

The United Soybean Research Retention policy will display final reports with the project once completed but working files will be purged after three years. And financial information after seven years. All pertinent information is in the final report or if you want more information, please contact the project lead at your state soybean organization or principal investigator listed on the project.