Client: Victorian Department of Environment, Land, Water and Planning (DELWP) [now Department of Energy, Environment and Climate Action (DEECA)]
Expertise: Machine Learning | Data Science | Project Management
The state-wide density product, with dense areas in purple, medium in light blue, and sparse in yellow
The challenge
Victoria’s vegetation data was nearly 20 years out of date and no longer accurate enough to support effective environmental analysis or decision-making. The Victorian Department of Environment, Land, Water and Planning (DELWP) is the owner and custodian of this data (which is part of Vicmap). DELWP needed a more automated, accurate and repeatable method of updating the data to increase operational efficiency, keep the data current, and improve resolution and quality for users. The goal was to apply machine learning to automate the update and maintenance of vegetation data, leveraging DELWP’s existing catalogue of imagery and elevation data.
The FrontierSI solution
FrontierSI worked closely with DELWP to first test and demonstrate the concept and performance of applying machine learning to existing Vicmap imagery across urban, peri-urban, rural, and forested areas. The methodology adopted to develop and test the machine learning algorithms and produce the final outputs was:
- Human analysts created training datasets over target areas.
- Developed a machine learning algorithm and compared the output against the training datasets
- Compared outputs to Vicmap specifications and requirements
- Delivered up-to-date, state-wide 20cm resolution tree extent and 2m resolution tree density data for public use.
The impact
The machine learning process and output vegetation data developed enable:
- Automation, increased efficiency, and significant cost savings for the Victorian Government in updating and maintaining this data.
- A repeatable and accurate process that ensures vegetation data is up-to-date and of higher quality.
- Smarter evidence-based environmental decision making.
This project shows how FrontierSI helps governments turn complex data challenges into practical solutions, supporting resilience for the environment and efficiencies for government.
Contact Us
Looking to automate the generation of vegetation or biodiversity datasets? Contact FrontierSI.