License Plate Recognition (LPR) technologies and airport parking solutions have a long rich history of usage together. These early adopters were challenged with costly deployments, requirements for additional large bright white lights to see the license plates, lane geometry that lent itself to post capture after a vehicle entered the facility, along with slow processing speeds and moderate LPR accuracy of the actual vehicle plate.
In addition to these early technology challenges were the need to accurately read all plates from all states at any one facility, stringent network requirements, deployments in an enterprise environment and the desire to move to virtual machines that offer failover and redundancy of the systems. These early deployments drove today’s innovative solutions for the airport parking industry.
Today’s LPR technology developers have solved most all of the earlier challenges by reducing costs more than 50 percent in the last decade. They have moved processing of the LPR image to the edge in the lane to increase speed, enhanced redundancy and minimized file size for improved network bandwidth, advanced deep learning neural networks, and increased the overall LPR read accuracy into a high 90 percent. The introduction of virtual loop technology allows the integrator to choose the most effective method of deployment between software and a hardware in ground loop.
Also, many airports, in cooperation with local law enforcement, implement a daily BOLO list in search of wanted vehicles and upon a BOLO hotlist match local authorities can be notified of a vehicle on the property.
The LPR systems in use currently offer features like touchless/frictionless ingress/egress for credentialed vehicles along with advance reservation holders. In addition, daily parkers/transients enjoy the benefits of LPR technology when they utilize the Pay on Foot kiosks deployed at the airport.
This is accomplished through traditional fixed lane LPR at the entry and exit plazas that link the parking event or ticket to the license plate. Upon exit, the plate is scanned and it is recognized in the PARCS system that the vehicle stay has been paid for and an automatic gate vend occurs for smooth, quick frictionless flow at the exit. In addition, many airports have moved to adding enhanced LPR capabilities by placing systems on vehicles for License Plate Inventory as well as handheld devices for smaller lot inventory needs.
By merging both the fixed LPR data with the mobile LPI data collected from the vehicles and handhelds, operators can now advise on occupancy per lot, look up lost ticket fees by license plate and, perhaps most importantly, they can offer “find my vehicle” services to enhance the customer experience. Kiosks can be placed at strategic locations at the property or in the parking office to assist customers who have misplaced their vehicle during the parking stay.
By simply entering at least one alpha numeric character from the license plate the individual can begin to hone in on the lot and row where they parked. All of the info was captured through the MLPI process performed by the airport, solving a very frustrating experience and getting the customer pointed in the correct direction to find their vehicle.
In addition, the MLPI process allows the facility to identify and locate vehicle overstays past a particular date like 30 days and/or find abandoned vehicles. All of these technology advancements have led to increased security of all vehicles, improvements of the parking experience, higher revenue generation and enhanced customer loyalty.
As our industry and technology continue to evolve, we are seeing greater demand for LPR processing on virtual machines that are configured for failover and redundancy. The use of software for vehicle detection is becoming more prevalent and eliminates the need for expensive in ground loops.
Also, several of us are working diligently to provide advance data sets of information from the vehicle which include vehicle make, model, state, color, speed, direction of travel, occupancy and classification. All of these data points assist in achieving the holy grail for LPR accuracy nearing 100 percent.
Imagine a red Toyota with a license plate of ABC 1234 and a trailer hitch obscuring one of the numbers. By matching to a set of data points, although one them may be off slightly, the certainty of matching to the correct vehicle for a gate vend is greatly improved.
John Whiteman is an expert in License Plate Recognition for over a decade with HTS which is now omniQ VISION. John can be reached at jwhiteman@omniQ.com