Weed managers guide to remote detection

Weed control is a costly and time-consuming issue for land managers across Australia. The detection of weed populations is always the first step towards achieving efficacious control. We are committed to investigating the possibilities and the limitations of remote sensing technologies for the detection of weeds in mixed landscapes.

The challenge

Remote sensing technology is currently used across various agricultural and environmental contexts but challenges remain. The expense and technical complexity of the technology often limits uptake by land managers, while complex natural systems make it difficult to distinguish weeds from other vegetation, especially during early stages of weed growth.

There is a need for those engaging in weed remote detection to access educational information and connect with others also working in this space to share learnings, resources and expertise.

Project name:
Weed managers guide to Remote Detection: Understanding the opportunities and limitations of multi-resolution and multi-modal technologies for remote detection of weeds in heterogenous landscapes (2021 - 2025)

Funding Support: Department of Agriculture, Water and Environment, $589,525

Our response

Our aim is to help land managers to effectively understand and use available and emerging remote sensing technologies for weed detection and provide a facility for them to access resources and connect with others in this field, thereby facilitating increased uptake and use of the technology to detect and manage weeds in various landscapes.

We are working with weed management experts from across agricultural and environmental fields to test the limits of current remote sensing technology for detecting weeds in diverse ecosystems, thereby improving mapping and monitoring of weed populations and achieving more effective on-ground weed management.

The result

We have developed a website to host the “The Weed Managers Guide to Remote Detection”, a set of guidelines reporting the learnings, outcomes and results of our recent project investigating the limitations and opportunities of this technology for detecting weed species in different environments. The website provides links to the newly established weed remote community of practice (WRS CoP), a community built to improve interactions between researchers, governments, natural resource management groups and a range of end-users to share learnings, provide advice and develop further research into remote weed detection. Pathways for accessing image libraries and models and a series of recent publications and case studies are also provided.

Access the Weed Managers Guide to Remote Detection Guidelines

In the media

MidCoast Council partner in research program using remote detection technologies to fight weeds

Gloucester Advocate

6 Aug 2021

Our team

Principal scientist

Dr Jane Kelly

janekelly@csu.edu.au View full profile

Our research team

Associate Professor Lihong Zheng

Data Science Research

View full profile

Dr Remy Dehaan

Spatial Scientist

View full profile

Associate Professor Felipe Gonzalez

Queensland University of Technology

Hillary Cherry

DPIE, Senior Weed Mgmt Coordinator, Weeds Team NPWS

Wendy Menz

DPIE

Mark Hamilton

DPIE

Key research publications

2023

Amarasingam, N., Hamilton, M., Kelly, J. E., Zheng, L., Sandino, J., Gonzalez, F., Dehaan, R. L., & Cherry, H. (2023). Autonomous detection of mouse-ear hawkweed using drones, multispectral imagery and supervised machine learning. Remote Sensing, 15(6), [1633]. https://doi.org/10.3390/rs15061633

2022

Zheng, L., Rahaman, M., Hamilton, M. A., Dehaan, R., Gonzalez, F., Kelly, J., & Cherry, H. (Accepted/In press). Remote Tiny Weeds Detection. In Lecture Notes in Computer Science: 10th Pacific-Rim Symposium, PSIVT 2022, Virtual Event, November 12–14, 2022, Proceedings (Vol. 13763).
Kelly, J., Rahaman, M., Sandino Mora, J., Zheng, L., Cherry, H., Hamilton, M. A., Dehaan, R., Gonzalez, F., Menz, W., & Grant, L. (2022). Weed managers guide to remote detection: Understanding opportunities and limitations of technologies for remote detection of weeds. In R. Melland, C. Brodie, J. Emms, L. Feuerherdt, S. Ivory, & S. Potter (Eds.), Proceedings of the 22nd Australasian Weeds Conference Adelaide September 2022 (pp. 58-62). Council of Australasian Weed Societies and the Weed Management Society of South Australia Inc.
Zheng, L., Oczkowski, A., Soomro, T. A., & Wu, H. (Accepted/In press). Rapid on-site weed identification with machine learning. In Image and Video Technology: 10th Pacific-Rim Symposium, PSIVT 2022, Bintan Island, Indonesia, November 12–14, 2022, Proceedings (Vol. 13763). (Lecture Notes in Computer Sciences; Vol. 13763). Springer.