Calculating Dogfennel (Eupatorium capillifolium) Control Using Machine Counting
Applied Research
J. Strickland
Extension Agent IV
UF/IFAS
Kissimmee
Abstract
Objective: Over the past three years, we have demonstrated the proof-of-concept of using machine counting to calculate percent control. This is the first fully replicated research project conducted using machine counting. Dogfennel (Eupatorium capillifolium) is a competitive pasture weed that reduces forage production. Traditional herbicide control assessments rely on subjective visual estimates, which may introduce variability. This study aimed to evaluate the use of machine counting as a standardized, objective method for assessing dogfennel control using aerial imagery and image analysis software. Methods: Nine 0.4-hectare plots were arranged in a completely randomized block design with three replications of each treatment: Weedmaster (3.52 L/ha), Pasturegard (1.75 L/ha), and an untreated control. Aerial RGB images were collected using a SenseFly eBee X drone and processed into an orthomosaic using Pix4D. The images were analyzed in ImageJ with a standardized color threshold (hue 41-89, saturation 127-255, brightness 0-94) to distinguish dogfennel from Bahiagrass (Paspalum notatum). Treatments were applied in June 2024, and control was evaluated in November 2024. Results: Untreated plots exhibited a 320.3% increase in dogfennel coverage. Pasturegard achieved 89% control, while Weedmaster provided 87% control. These findings closely matched visual assessments, validating the accuracy of machine counting. Conclusion: Machine counting using ImageJ software proved to be a reliable tool for assessing herbicide effectiveness. This method offers a more objective and replicable alternative to traditional visual estimates. AI-driven analysis can enhance Extension research, providing precise, data-driven recommendations for weed management. Future work should refine AI integration to further improve herbicide evaluation methodologies.
Poster has NOT been presented at any previous NACAA AM/PIC
This poster is being submitted for judging. It will be displayed at the AM/PIC if not selected as a State winner. The abstract will be published in the proceedings.
Click to view Poster
Authors: J. Stacy Strickland, Jessica Sullivan, J.J. White
-
Strickland, J. Extension Agent IV, UF/IFAS, Florida, 34744
-
Sullivan, J. Extension Agent IV, UF/IFAS, Florida, 34744
-
White, J. Soil and Water Conservation, Osceola County, Florida, 34744