Skip to main content

Q-Free neural networks see all sides

Analytics solution extends Intrada ALPR suite
By David Arminas June 2, 2020 Read time: 2 mins
A black and white case: Q-Free also identifies vehicle class, colour, make and model (© Skovalsky | Dreamstime.com)

Q-Free says it is developing improved vehicle analytics and detection for its automatic licence plate recognition (ALPR) technology that identifies vehicle class, colour, make and model.

The technology also identifies which side of the vehicle is being analysed, according to Q-Free.

The vehicle analytics feature is an extension of the Norway-based company’s Intrada ALPR which processes more than a billion licence plates around the globe each day.

The solution can be used with any vendor, making it a convenient extension that opens new possibilities for its customers’ operations and business models.

There is no need for a customer to change existing video infrastructure or invest in costly hardware-based alternatives such as radar and laser, says Q-Free.

The company says that data from test sites in South America and Asia show surveillance and security operators successfully gathering additional identifying characteristics to make the best use of existing video detection equipment.

In particular, the vehicle angle feature determines which side of the vehicle is facing the camera, for example the front or rear. This is helpful in determining entry and exit points in parking applications.

The new vehicle analytics are a result of innovative, reliable neural networks and the company’s machine learning capabilities, according to Marco Sinnema, product manager for Q-Free’s Intrada ALPR library.

“Work with initial customers continues to train the detection of the neural networks – which is now available in our commercial, off-the-shelf Intrada ALPR library,” he said.

“Early results are showing the system performing with great precision, and we plan on delivering the same unrivalled automation accuracy and low error rates offered in our existing ALPR solutions.”

Q-Free’s other products and brands include Intelight, OpenTMS, Intrada, ParQSense and Q-Free Hub.

For more information on companies in this article

Related Content

  • Why keeping count is so important for traffic management
    November 21, 2023
    Traffic engineers need to have multiple solutions in their toolbox to complete the most accurate and safe data collection programmes possible, explains Wes Guckert of The Traffic Group
  • Australian tolling industry debates tag replacement
    June 11, 2014
    Australia’s 2014 National Electronic Tolling Conference (NeTC) inspired lively debate among the 130 delegates about tackling the need to replace seven-plus million tolling tags that are reaching the end of their life. In his opening address, Australian Toll Road Users’ Group Chair Rex Wright said the industry was potentially facing a US$94 million bill over the next five years to replace old tags. As Australian tolling authorities operate a harmonious national tag system, all toll operators are committed
  • Rekor develops Rekor Edge vehicle recognition system
    August 20, 2019
    Rekor Systems is to launch a camera recognition solution which it says is designed to read the vehicle make, model, number plate, colour and body type. The Rekor Edge solution comes with an accuracy rate exceeding 99% and is suitable for parking and law enforcement, the company adds. Rod Hillman, chief operating officer of Rekor, says: “With Rekor Edge, we’re democratising vehicle recognition solutions by bringing a cost-efficient and user-friendly solution to potential markets.” According to Rekor, the
  • Xerox passenger detection system solves HOV and HOT enforcement challenge
    July 31, 2015
    High Occupancy Vehicle (HOV) lanes and High Occupancy Toll (HOT) lanes are powerful ITS tools. However, the challenge of how to enforce them is proving a constant barrier. Manual enforcement is both expensive (dedicated police resource) and inefficient (poor detection rates). Xerox estimates that the ability to identify violators in the HOV/ HOT lanes by use of human eyesight is less than 20%.