ALPR has become an important and necessary tool for law enforcement as well as other end user organizations. In this article we discuss what’s new with ALPR and why we believe growth is likely to continue.
We’ve discussed automatic license plate recognition (ALPR) in
great length. Indeed it has become an important and necessary tool for law enforcement as well as other end user organizations. In this article we discuss what’s new with ALPR and why we believe the technology is here to stay.
ALPR has gained momentum among various end user entities including cities, municipalities, law enforcement, and private parking companies.
Demand for this technology has remained strong. According to MarketsandMarkets, the global ALPR system market is projected to reach US$4.8 billion by 2027, registering a CAGR of 9.2 percent from 2022 to 2027. Below are some of the reasons why we believe ALPR is expected to remain popular and see continued growth.
Smarter and more intelligent
ALPR is by no means a new technology and has been here for two decades. But advances in AI and deep learning have made it smarter and more accurate than ever.
“At Genetec, we use deep learning in our ALPR solution. We're training our algorithms using a structured dataset of raw LPR images and a limited set of possible classes or outputs. Structured datasets consist of data that has been organized or labeled in a predefined manner. In this case, the datasets include labeled images of a wide variety of license plates,” said Larry Legere, Commercial Director for AutoVu at Genetec. “The goal is to have the system take an image of the rear of a car that it has never seen before and be able to output the license plate characters, its state of origin, and the make of the vehicle.”
“Recent updates to ALPR go beyond license plate recognition, allowing for the identification of the make, model, color, identifying marks, state, and direction of travel of vehicles. In fact, this is further expanded to unique characteristics that a vehicle may have,” said Charles Degliomini, Executive VP at Rekor Systems. “For example, a vehicle of interest may have a roof rack, or have significant rust on a panel. The AI engine gathers this information and can utilize it to better search for vehicles that meet these unique characteristics.”
More diverse applications
ALPR now has more diverse applications. Security-wise, helping officers detect problematic or suspicious vehicles on the spot has become quite common. Beyond security, ALPR deployed at parking lots boosts the user experience. Further, ALPR provides valuable vehicle- and traffic-related data that users can refer to for planning and management purposes.
“In addition to the usual law enforcement use cases, ALPR can offer traffic management agencies some very useful information without compromising on the privacy of citizens. Most ALPR cameras have the ability to categorize the vehicles and measure their approximate speeds. They can be used as traffic sensors to give average speeds, traffic volume and vehicle classification. Since the cameras can uniquely identify each vehicle, they also make for great sensors for travel times, wait times and origin-destination studies,” Legere said.
Rekor, meanwhile, cited certain unique use cases, including electric vehicle identification and environmental impact tracking, that make cities more environmental and sustainable.
“Understanding patterns and volumes of electric vehicles, hot spots, and changes over time may allow agencies to better perceive where to place electric vehicle charging stations, in what quantity, and how to stay ahead of the rapid adoption of all-electric vehicles in the US,” Degliomini said.
He added: “GHG emissions measurements are an important way to better understand the air pollution occurring in a city and in developing strategies to help curb the global environmental impact. Sensors capturing vehicle model and make can also press this information against GHG emission estimates, providing aggregate estimated concentrations of GHG emissions across different locations in cities. This can be used to better understand where, when, and how these pollutants are entering into the environment, and to help determine strategies that alleviate pollution.”
More flexibility in deployment
ALPR can be deployed both on server and on the
edge. While server provides more computing power, more and more
ALPR cameras are now available to process images closer to the data point. A hybrid architecture is also available, combining the best of both worlds.
“ALPR can be captured through cameras mounted at key intersections and highways to identify and get alerted on vehicles of interest. For mobile patrols, high-priority vehicle data can be captured through in-vehicle cameras, or smartphone apps from investigators on foot. ALPR algorithms can also easily be put on servers that allow customers to leverage existing infrastructure more readily and get much more value out of what they have already put in place,” Degliomini said. “For the most comprehensive view of the roadway in a real-time, scalable, and affordable way, it’s most optimal to deploy the latest in edge computing camera systems out on the side of the road, where vehicle recognition can be done at point of capture, and metadata can be sent up that requires much less bandwidth.”
“Genetec is now offering a new solar based camera that is a hybrid approach. The camera features machine learning DNN on the camera to detect certain vehicle information and then sends the vehicle images to a cloud service to process additional information like vehicle type, make, model and color,” Legere said.