RMIT geospatial scientists build AI tool to monitor street infrastructure

By on 19 June, 2019

A team led by a geospatial science honours student at RMIT has developed a fully automatic system to monitor street signs needing repair that runs off Google Street View imagery.

The team recently published their results in the Computers, Environment and Urban Systems journal, reporting that their system returns 96 percent accuracy in recognising signs, identifying them with 98 percent accuracy, then records their geolocation from the available 2D imagery.

The team says that their system could lead to considerable cost savings for authorities, and its uptake would help minimise risks to workers over a workflow of manual inspection.

Andrew Campbell, lead author of the study, said that the tool was scalable and easily extensible with new data sources, with their proof-of-concept focusing on Australian ‘Stop’ and ‘Give Way’ traffic signs.

“Councils have requirements to monitor this infrastructure but currently no cheap or efficient way to do so. By using free and open source tools, we’ve now developed a fully automated system for doing that job, and doing it more accurately,” he said.

The research team reported that the mandatory database for sign locations contained inaccuracies of up to 10 metres.

“Tracking these signs manually by people who may not be trained geoscientists introduces human error into the database. Our system, once set up, can be used by any spatial analyst – you just tell the system which area you want to monitor and it looks after it for you,” Campbell said.

Campbell credited the project’s initial concept to his industry mentor at Alpine Shire Council and RMIT Geospatial Science alumnus, Barrett Higman. The study was co-led by Dr. Chayn Sun and fellow RMIT geospatial scientist Dr Alan Both, from the university’s Centre for Urban Research.

Stay up to date by getting stories like this delivered to your mailbox.
Sign up to receive our free weekly Spatial Source newsletter.

You may also like to read:


, , , , , ,


Newsletter

Sign up now to stay up to date about all the news from Spatial Source. You will get a newsletter every week with the latest news.

  • This field is for validation purposes and should be left unchanged.
Mapping invasive grasses with machine learning
Researchers have developed a model that can identify Gamba g...
Leica launches pint-sized smart antenna, Zeno FLX100
Leica lobs a tiny, modular new GNSS antenna for field captur...
Q&A with Andy Sexton
We sit down with Andy Sexton, whose Navigat/AR system seeks ...
Record for largest drone swarm broken twice in September
Mind-bendingly large and complex aerial performances from UA...
Faro launches enterprise edition of Webshare suite
Faro has launched an enterprise edition of its Webshare 3D r...
Nextcore launches RN100 UAV LiDAR unit
Nextcore launches its next-generation LiDAR sensor for UAVs....