Related Experiment Video
Updated: Oct 7, 2025
14:44
Harmonic Radar Tags for Insect Tracking: Lightweight, Low-cost, and Accessible
Published on: May 13, 2025
1.3K
OpenWeedLocator (OWL): an open-source, low-cost device for fallow weed detection.
Guy Coleman1, William Salter2, Michael Walsh2
1School of Life and Environmental Sciences, Sydney Institute of Agriculture, The University of Sydney, Brownlow Hill, NSW, Australia. guy.coleman@sydney.edu.au.
Scientific Reports
|January 8, 2022
Summary
OpenWeedLocator (OWL) is a new, low-cost, open-source device for detecting fallow weeds using image analysis. This technology enhances site-specific weed control, making it more accessible to the agricultural community.
Area of Science:
- Agricultural Science
- Computer Vision
- Precision Agriculture
Background:
- Fallow phases are crucial for crop yield in dry environments, requiring effective weed management.
- Current site-specific weed control relies on expensive, proprietary sensor systems.
- Image analysis offers potential for developing accessible weed detection technology.
Purpose of the Study:
- To introduce OpenWeedLocator (OWL), an open-source, low-cost, image-based device for fallow weed detection.
- To promote community engagement in developing site-specific weed control methods.
- To validate the effectiveness of OWL for weed detection in agricultural settings.
Main Methods:
- Development of an open-source, image-based device (OWL).
- Utilizing four established color-based algorithms for weed detection.
- Field validation across seven fallow fields in New South Wales, Australia.
Main Results:
- OWL demonstrated effectiveness in detecting fallow weeds.
- Average precision of 79% and recall of 52% were achieved across fields.
- Individual transects showed high performance, with up to 92% precision and 74% recall.
Conclusions:
- OpenWeedLocator (OWL) provides a low-cost, accessible solution for fallow weed detection.
- The open-source approach fosters community-driven development in precision agriculture.
- OWL has the potential to improve weed management strategies and enable in-crop applications.
