commons/openweedlocator
Low-cost open-source weed detection device for site-specific spraying, published with printable enclosures and camera housings for several build variants.
OpenWeedLocator (OWL)
Open-source, low-cost weed detection for site-specific weed control
OWL is a camera-based weed detection system based on the Raspberry Pi that uses green-detection algorithms to trigger relay-controlled solenoids for spot spraying. It's built entirely from off-the-shelf components and 3D-printable parts, making precision weed control accessible to anyone.
Website | Documentation | Community | Newsletter
OWLs in Action
<table> <tr> <th>2m vehicle OWL</th> <th>2m robot-mounted OWL</th> <th>Bicycle OWL</th> </tr> <tr> <td align="center"> <img src="https://user-images.githubusercontent.com/51358498/130522810-bb19e6ca-5019-4de4-83cc-858eca358ef8.jpg" width="250"/> </td> <td align="center"> <img src="https://github.com/geezacoleman/OpenWeedLocator/assets/51358498/9cb73514-dffc-4c53-969e-c1c816610f1b" width="250"/> </td> <td align="center"> <img src="https://github.com/geezacoleman/OpenWeedLocator/assets/51358498/17ad4ead-429e-4384-9e74-b050a536897f" width="250"/> </td> </tr> <tr> <th>12m X-fold OWL (in development)</th> <th>4m OWL sprayer</th> <th>16 channel vegetables OWL</th> </tr> <tr> <td align="center"> <img src="https://github.com/user-attachments/assets/c39308cf-ccd7-4428-a10f-fef2fa0ae5af" width="250"/> </td> <td align="center"> <img src="https://github.com/user-attachments/assets/f32c307d-7fa4-4907-82f2-53519e2236b8" width="250"/> </td> <td align="center"> <img src="https://github.com/user-attachments/assets/b8c8027a-bacd-41e3-9595-8786ff16b715" width="250"/> </td> </tr> </table>Quick Start
# Clone and install on Raspberry Pi (Bookworm or Trixie OS)
git clone https://github.com/geezacoleman/OpenWeedLocator owl
bash owl/owl_setup.sh
During this process you'll be asked to setup:
- Green-on-Green - this adds about 2GB of dependencies
- Dashboard - standlone or networked
One complete you'll need to reboot and then it should be running.
To confirm, run sudo systemctl status owl.service or sudo journalctl -u owl.service -f
See the two step installation guide for detailed instructions.
Documentation
- Getting Started - What you need, how it works
- Hardware Assembly - Parts list, wiring, 3D printing
- Software Setup - Installation, configuration, algorithms
- Controller Setup - Standalone and networked dashboards
- Troubleshooting - Common issues and fixes
Contributing
Contributions are welcome! See CONTRIBUTING.md for guidelines, or join the conversation at community.openweedlocator.org.
Citing OWL
OpenWeedLocator was originally published in Scientific Reports.
@article{Coleman2022,
author = {Coleman, Guy and Salter, William and Walsh, Michael},
doi = {10.1038/s41598-021-03858-9},
issn = {2045-2322},
journal = {Scientific Reports},
number = {1},
pages = {170},
title = {{OpenWeedLocator (OWL): an open-source, low-cost device for fallow weed detection}},
url = {https://doi.org/10.1038/s41598-021-03858-9},
volume = {12},
year = {2022}
}
The OWL speed testing paper has been published in Computers and Electronics in Agriculture. Please consider citing the published article using the details below.
@article{Coleman2023,
author = {Coleman, Guy R.Y. and Macintyre, Angus and Walsh, Michael J. and Salter, William T.},
doi = {10.1016/j.compag.2023.108419},
issn = {0168-1699},
journal = {Computers and Electronics in Agriculture},
pages = {108419},
title = {{Investigating image-based fallow weed detection performance on Raphanus sativus and Avena sativa at speeds up to 30 km h$^{-1}$}},
url = {https://doi.org/10.1016/j.compag.2023.108419},
volume = {215},
year = {2023}
}
License
This project is licensed under the MIT License.