Allows use of up to twice as many cameras while reducing computing cost
Rekor Systems, Inc. (NASDAQ: REKR) ("Rekor"), a leading provider of innovative vehicle recognition systems, announced today that it has released an upgrade to its OpenALPR vehicle recognition software. The newly released software improves processing speed on Central Processing Units (“CPU”) up to 100%, thus enabling customers to reduce costs by deploying up to twice as many cameras using the same computing hardware.
Other upgrades include performance improvements on Nvidia Graphics Processing Units (“GPU”), which decrease memory usage on both CPU and GPUs and reduce reads of road signs and billboards by mobile units. Already operating at a precision 99.02% accuracy rating on U.S. license plates, OpenALPR was able to improve accuracy across all vehicle recognition metrics, based on publicly available benchmarks.
The new software also includes enhancements that improve user experience, including a new dispatch view for real-time alerting in map-view, text message alerts, usability enhancements for analytic reports, and improvements in overall efficiency.
“We’ve already exceeded human-level accuracy benchmarks, so to be able to develop an even more accurate solution is a giant leap for vehicle recognition technology,” said Matthew Hill, Chief Science Officer, Rekor. “Doubling processing speed can be a game-changer for customers who want to increase the number of vehicle recognition cameras they’re running. Now they’ll be able to do so without incurring additional capital expense for the purchase of supporting hardware. This represents a significant cost reduction while improving overall camera coverage. Our customers rely on OpenALPR for its speed, accuracy and scalability, so we’re pleased that this enhanced version will also reduce their need to purchase additional expensive hardware.”
With its robust and growing license plate database covering 69 countries, OpenALPR's software can identify in real time vehicle license plate data, color, make, model and body type.