Automated detection of road cracks
CSIRO’s RoadCrack imaging technology automates the task of surveying the nation’s roads with a capacity to detect cracks one millimetre wide from beneath the chassis of a moving vehicle.
- 19 August 2010 | Updated 14 October 2011
CSIRO's RoadCrack, developed in collaboration with the NSW Roads and Traffic Authority (RTA), is a mobile camera and image-processing tool that detects, classifies and reports on cracking in road pavement.
RoadCrack is the first system in the world to achieve the RTA's standards for an automated, objective and accurate crack detection system.
It has successfully surveyed hundreds of thousands of kilometres of Australian roads.
The problem
Manual surveying the condition of road surfaces suffers problems of:
- cost
- accuracy
- safety
- efficiency.
The RTA approached CSIRO to assess the potential of automated crack detection technologies.
What CSIRO did
Our solution, RoadCrack, is a fully-automated system.
It combines advances in:
- machine vision
- parallel computing
- artificial intelligence
- image analysis.
High-speed cameras are mounted under a vehicle to collect digital images of the pavement surface while the vehicle is moving at highway speeds.
A special reflector system focuses light to illuminate tiny cracks.
High-resolution images are collected for small sections of pavement and then consolidated into bigger images that cover half-metre intervals of the road surface.
CSIRO also developed algorithms and computer hardware for extracting the relevant information in real time.
| | |
During operation, road images are passed to a processing module that uses sophisticated image analysis algorithms to classify cracks as they are detected.
Cracks are reported according to type, severity and extent.
Outcomes
RoadCrack detects cracks as fine as one millimetre in width, while travelling up to 105 kilometres per hour.
It can also operate at night as the system provides its own lighting for road imaging.
The objective data from RoadCrack provides valuable input into management of road pavement assets, saving the NSW State government roads authority tens of millions of dollars each year in road maintenance costs.
Read about Safe-T-Cam: keeping an eye on the road.
Commercial Information
Challenge: to improve monitoring and maintenance of road surfaces
Solution: revolutionary mobile image-capture tool that combines advances in machine vision, parallel computing and artificial intelligence
Opportunities: widespread application in transport industry
CSIRO Divisions:
- Materials Science and Engineering
- Forest Biosciences
- Mathematics, Informatics and Statistics
Partners: NSW Roads and Traffic Authority