Skip to main content

Kapsch upgrades North America ANPR engine

It will ID licence plates from 58 jurisdictions in US, Mexico and Canada
By Adam Hill February 27, 2025 Read time: 2 mins
Over 150 plate types can now be accurately identified (© Swisshippo | Dreamstime.com)

Kapsch TrafficCom has upgraded its automatic number plate recognition (ANPR) engine for the US and Canada as part of its tolling offering, with deployment in several projects underway.

The company introduced regional ANPR engines in Europe and Australia in 2023.

“A newly-developed deep neural network - specifically for North America - ensures precise identification of licence plates from 58 jurisdictions, including all mainland US states, Hawaii, Mexico, and the most relevant Canadian provinces,” explains Chris Jones, director of video transaction optimisation at Kapsch TrafficCom North America. 

Unlike traditional project-specific engines, the new product "is designed to serve multiple projects across the region, offering unparalleled accuracy and reliability".

JB Kendrick, president at Kapsch TrafficCom North America, adds: “Our newest and most robust engine can accurately identify over 150 plate types, focusing on the most widely used designs and ensuring easy scalability to include additional types as needed."

It will reduce manual review costs by improving the percentage of plates read automatically and represents "a key building block for our existing and future roadside tolling clients", Kendrick says.

The company says North America presents "unique challenges" for ANPR systems because of the sheer variety of licence plate designs, syntaxes and allowed characters. The new ANPR engine has been "meticulously trained" to handle this complexity, the manufacturer insists.

For more information on companies in this article

Related Content

  • Jenoptik showcases global family of traffic enforcement and monitoring
    March 1, 2016
    Jenoptik Traffic Solutions says it will use Intertraffic Amsterdam 2016 as the perfect platform to present its global family of enforcement and monitoring technologies. The company says that with over 30,000 delivered systems, operating in more than 80 countries around the world, it understands very well that not all customers have the same requirements; one size does not fit all.
  • Making the case for ALPR in enforcement
    February 2, 2012
    Federal Signal's Brian Shockley uses examples from around the world to make the case for the greater use of automatic license plate recognition technology in the US. It is time, he says, to consider the possibilities of a national network and the use of average speed enforcement
  • Hikvision adds AI to DeepinView cameras
    July 22, 2020

    Hikvision has added to its DeepinView camera line with what it calls the Dedicated subseries.

    The company says this loads a batch of artificial intelligence-powered deep learning algorithms into each unit, which are of use in two distinct categories.

    The first is vehicle analysis, where cameras combine automatic number plate recognition with vehicle attribute recognition - including make, colour and direction of movement. 

    Typical installations would be on city streets and at the entrances of car parks and buildings.  

  • Machine vision makes progress in traffic applications
    June 2, 2014
    Machine Vision technology is easing the burden on hard-pressed control room staff and overloaded communications networks.