Software developed by a team at the University of Maryland at College Park makes use of crowdsourcing to report road problems to local governments.
Untrained crowdsourced workers from Amazon's Mechanical Turk (AMT) service can use Google Street View to search streets and label roads that have potholes, obstructed pathways, broken curbs, or missing ramps. The labeling software can generate a report that includes images of highlighted problem areas.
Officials at the U.S. Department of Transportation have expressed interest in the system and believe it can save local governments time and resources on road repairs by telling repair teams the precise nature of the problem before they are dispatched into the field.
In a test, users who viewed a short instructional video accurately spotted access issues 93 percent of the time. The researchers want to automate the process using computer-vision algorithms for the next version of the system. They also want to increase the accuracy of the system by accessing the precise laser range-finding data the Street View camera cars acquire as they drive along the streets.
From New Scientist
View Full Article
Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA