Enhancing Security and Business Intelligence

Modern ALPR software running on the edge

Automated License Plate Recognition (ALPR) has transformed significantly over the years, evolving from a niche technology into a powerful tool for a wide range of applications, particularly in border security.

From border security to parking lots, ALPR has gained traction across multiple use cases as the technology becomes more accurate and affordable than ever. I spoke with Jason Cook, business development director at Vaxtor, a leader in ALPR AI-based analytics, and Rui Barbosa, category manager, Surveillance Products at i-PRO, a maker of AI-enabled security cameras, to delve into the latest advancements and applications of ALPR technology.

Q. How has ALPR technology evolved over the years, and what are the main factors driving this change?

Jason Cook: ALPR technology has come a long way from its early days, when it required specialized, bulky cameras costing thousands of dollars. Today, advancements in image quality and processing power of modern AI-enabled IP cameras have democratized ALPR, making it more accessible and cost-effective.

The powerful chipsets within contemporary CCTV cameras allow ALPR software to be hosted solely within the camera as an edge process. This shift means that standard surveillance cameras, costing only a few hundred dollars, can now perform high-accuracy license plate recognition tasks that previously required specialized equipment.

Rui Barbosa: The integration of ALPR software into regular CCTV cameras has several benefits. It simplifies installation, reduces costs, and maintains the primary function of surveillance while adding the powerful feature of license plate recognition.

This development has made ALPR more versatile and widely applicable, particularly in scenarios where budget constraints and existing infrastructure need to be considered.

Q. How accurate is ALPR today, and what additional capabilities are possible with modern systems?

Jason Cook: One of the significant advancements in ALPR technology is its remarkable accuracy. At Vaxtor, we achieve around 99% accuracy rates in most scenarios in over a hundred countries with detection rates above 99%. Such high accuracy is crucial for border security applications, where reliable vehicle identification is essential for maintaining security and operational efficiency.

Beyond license plate recognition, modern ALPR systems also offer vehicle classification, motion direction, make, model and color recognition capabilities. It also can accurately recognize the country origin of the license plate in its province in some cases (USA, Oceania, Egypt, Thailand, among others).

This is particularly useful in law enforcement and border security, where identifying vehicle characteristics can aid in tracking and monitoring suspect vehicles. Our software can recognize more than 99% of the vehicle models circulating in the streets, ensuring accurate identification even in diverse geographical areas.

Q. How does edge computing benefit ALPR technology, especially in high-security environments?

Rui Barbosa: A significant trend in ALPR technology is the move toward edge computing. Instead of relying on powerful backend servers, modern ALPR systems can process data on the edge directly within the camera. This approach not only reduces costs but also simplifies installation and maintenance. Even existing cameras can be retrofitted with ALPR capabilities using lightweight software that runs efficiently on minimal hardware.

Processing ALPR on the edge offers several advantages in many use cases. It reduces the need for extensive data transmission, ensuring faster response times and lower latency. This is particularly important in high-security environments where timely data processing is critical. Additionally, edge-based ALPR systems can operate independently of network connectivity, providing reliable performance even in remote or challenging locations.

Q. How important is collaboration between ALPR software providers and camera manufacturers?

Jason Cook: The collaboration between ALPR software providers and camera manufacturers is crucial for optimizing performance. Our partnership with i-PRO exemplifies this synergy. By working closely with camera manufacturers, ALPR providers can ensure that their software leverages the full potential of the camera hardware. For instance, specific lens choices and camera positioning can significantly impact the accuracy and effectiveness of ALPR systems.

Rui Barbosa: Our latest AI-enabled cameras use power SoCs, such as Ambarella’s CV-52. These powerful processors can host multiple simultaneous applications such as VaxALPR and more.

The lens and camera placement choice are vital for high-speed scenarios, such as highways. Cameras must capture clear images of license plates from various angles and distances. This collaboration and attention to detail enhances the reliability of ALPR systems for every application.

Q. How do modern ALPR systems adapt to environmental challenges like lighting and weather conditions?

Jason Cook: Environmental factors such as lighting conditions and weather can affect the performance of ALPR systems. Modern ALPR technology addresses these challenges through advanced image processing and adaptive algorithms. Our software can manage varying lighting conditions, ensuring accurate plate recognition even in low light or harsh weather.

Rui Barbosa: Both the camera and the software’s adaptability are crucial, as vehicles may pass through at any time of day and in different weather conditions. Having adequate lighting, whether through IR illumination from the camera or additional ambient lighting, ensures that ALPR systems perform consistently regardless of external factors.

Q. What are some expanding use cases for ALPR technology beyond border security, law enforcement, and parking management?

Jason Cook: While ALPR technology is widely used in law enforcement and parking management, its applications are expanding into other areas. In border security, ALPR systems can monitor and control vehicle access, track suspect vehicles, and gather data for intelligence and analysis. Additionally, ALPR is increasingly used in retail environments to collect demographic data, monitor customer behavior, and enhance operational efficiency.

Rui Barbosa: Access control via ALPR has grown significantly as prices for systems have decreased. For example, some retailers offer opt-in recognition for preferred customers to offer specialized services. Schools are also rolling out ALPR as a frictionless access control method for students and staff.

Q. How do privacy and regulatory considerations affect the deployment of ALPR systems?

Jason Cook: As ALPR technology evolves, so do privacy and regulatory compliance considerations. In regions like Europe, stringent regulations such as GDPR require careful handling of personal data. ALPR providers must ensure that their systems comply with these regulations, protecting the privacy of individuals while providing robust security solutions. Rui Barbosa: I some countries, ALPR-enabled cameras on private property may only be allowed to capture plates entering the property. Cameras must be mounted at distinct angles and in some cases, masking is employed to limit the cameras range of view. Ensuring transparency and adhering to regulatory requirements helps maintain public trust and supports the ethical use of ALPR technology.

Q. In conclusion, how do you see the future of ALPR technology advancing?

Jason Cook: The advancements in ALPR technology, driven by improvements in camera hardware and image processing capabilities, have made it a powerful tool for various applications. The integration of ALPR software into standard CCTV cameras, the move towards edge computing, and the collaboration between ALPR providers and camera manufacturers have significantly enhanced the accuracy, efficiency, and accessibility of ALPR systems.

We are getting requests from organizations wanting to enhance the automation of tasks, which requires more general optical character recognition (OCR). Our latest VaxOCR Genesis reader is an example of that technology that recognizes Latin characters and numbers, such as ID cards, stock labels, receipts and more.

Rui Barbosa: As AI-enabled cameras become increasingly sophisticated, we expect to see more emphasis on accurately detecting make, model, and color information in addition to plate numbers to recognize vehicles of interest and enhance verification of vehicles customers wish to track. When integrated with AI-based analytics, ALPR and generic OCR provide a wealth of additional information. Whether it is parking, law enforcement, traffic management, security, access control, logistics or smart cities, advances in ALPR technology are paving the way for streamlined security and operations.

This article originally appeared in the September / October 2024 issue of Security Today.

Featured

Featured Cybersecurity

Webinars

New Products

  • HD2055 Modular Barricade

    Delta Scientific’s electric HD2055 modular shallow foundation barricade is tested to ASTM M50/P1 with negative penetration from the vehicle upon impact. With a shallow foundation of only 24 inches, the HD2055 can be installed without worrying about buried power lines and other below grade obstructions. The modular make-up of the barrier also allows you to cover wider roadways by adding additional modules to the system. The HD2055 boasts an Emergency Fast Operation of 1.5 seconds giving the guard ample time to deploy under a high threat situation. 3

  • 4K Video Decoder

    3xLOGIC’s VH-DECODER-4K is perfect for use in organizations of all sizes in diverse vertical sectors such as retail, leisure and hospitality, education and commercial premises. 3

  • EasyGate SPT and SPD

    EasyGate SPT SPD

    Security solutions do not have to be ordinary, let alone unattractive. Having renewed their best-selling speed gates, Cominfo has once again demonstrated their Art of Security philosophy in practice — and confirmed their position as an industry-leading manufacturers of premium speed gates and turnstiles. 3