Adaptive Recognition on Advancing ANPR Through Continuous Innovation

Watch here

In this edition of In the Spotlight by Parking Network, we are joined once again by Sam Benkoczy from Adaptive Recognition. With decades of experience in vehicle recognition, from image processing and ANPR to broader vehicle analysis, Adaptive Recognition has built its approach around delivering reliable vehicle data. In this interview, we explore how the company supports partners in increasingly complex environments, simplifies integration, and continues to evolve its solutions with new functionalities and customer-driven developments.

Q: Could you briefly reintroduce the company?

Sam Benkoczy:

Here at Adaptive Recognition, our roots are really deep in image processing. When we see the image of a vehicle, we get excited and we want to extract all the possible data from it and give it to you. That has remained our core mission ever since. We want to provide reliable vehicle data to our partners.

It all started with ANPR and licence plate recognition back in the 90s. Then, as technology moved forward, we moved into make and model recognition, and more recently, we are into full vehicle analysis. We started as a software company, but then we realised that to get the best results out of the images, we had to build our own hardware, because that hardware can provide an optimised image for the kind of analysis that we want to do.

We are not in a race for megapixels, unlike in the cell phone world or the photography world. We are focused on giving the best-quality images for vehicle recognition.

Q: Where do you see the main complexity in the industry today?

Sam Benkoczy:

Things are becoming interconnected. The industry is no longer separate islands, but there is a lot of integration happening. Our specific challenge is not so much to see the complexity of everything, but we have always remained focused on that mission. Even amidst this complexity, we want to give reliable data, because that is the foundation of everything.

That is why we have our own factory.

Specific challenges are always dependent on location. In some places, it is the lack of front plates, so you have to read from the back. In other places, it is the weather conditions that do not allow for perfect images all the time because they get mucky.

Then, people in certain cultures are more into tailgating than others. Then we have fraudulent plates, where people want to trick the system with a printed image. Then there is the ever-present challenge of network issues.

So we try to come up with creative ways and very smart ways to give the best results despite these challenges.

Q: Adaptive Recognition seems to focus on making things simpler for partners and customers. How do you approach that?

Sam Benkoczy:

We have many partners, and we want to aim at keeping their lives really simple and making integration simple.

We have a number of native integrations with third-party systems, as well as our API being readily available, with plenty of sample code.

When we have a new customer, we have an onboarding call and we roll out the red carpet. We help them make sure that they have all the tools necessary to make that integration as seamless and as fast as possible.

Q: Would you say this ease of integration is one of the things that sets Adaptive Recognition apart?

Sam Benkoczy:

Yes, that is one of the things, but I could go on. The list goes on and on.

We have a dedicated team of over 100 people in research and development, and they work tirelessly to improve the experience, improve accuracy and speed, and include new makes and models. Some of the vehicles that we see on the roads today did not exist five or ten years ago. It is an ever-evolving world.

We give firmware releases on a regular basis, and we always try to make sure that you have a familiar camera with new added functionalities to enhance your system. So if you compare our cameras today to the cameras that we had two years ago, although it is the same hardware, you are comparing apples to oranges.

We also aim to make sure that you take the camera out of the box and have it ready and up and running within five minutes. It does not take very long to fine-tune them either. You can have them up and running in five minutes, and then in another five minutes, you can have them running perfectly and perfectly dialled in to your environment.

Our development is always based on customer feedback and customer need. That is also very important. We joke that our strongest competitors are the cameras that we made ten years ago, because those are so good in quality that they refuse to die.

Q: Could you give us a few examples of the improvements or new functionalities that have been added to existing products over time?

Sam Benkoczy:

For sure. We have ANPR and make and model recognition on our cameras, but then people wanted to know which direction the vehicle was moving. So we have direction detection.

People wanted to have cameras trigger-less out of the box. Not everybody wants to pay a lot of money and go through the pain of putting in induction loops. Then we can talk about reading on multiple lanes and the native integrations that we introduced.

We had a request for tampering detection. If somebody covers the camera or pushes the camera, those are functions and features that we introduced recently.

Also, as I mentioned, fraud detection.

It is interesting because in some regions, people were trying to trick the cameras by showing them images of a licence plate, printed images, or a photograph they took of a vehicle. More recently, we introduced something called fraud detection.

So if somebody wants to trick the camera that way, we give a score: how likely is it that the camera is seeing a real vehicle with a real licence plate? These are all functions that we have been working on.

Q: What does the future look like for ANPR and vehicle recognition? And how is Adaptive Recognition preparing for that next phase?

Sam Benkoczy:

Moving forward, we continue to listen to our customers and their feedback. We continue to focus on what is important to them and where the market is leading us.

Be it new functions in the same camera, new functions involving multiple cameras, tailgating detection, front and rear plate matching, or real-time parking enforcement, anywhere that the road leads us.

There are a lot of interesting features and interesting things that the future holds.

About Adaptive Recognition

Adaptive Recognition is a global leader in access control and intelligent recognition, delivering advanced technology to enhance security, simplify operations, and improve efficiency. AR's innovative hardware and software integrate seamlessly with various parking platforms, offering a flexible and user-friendly approach to access control and parking enforcement.

AR specializes in designing and manufacturing components that support the installation of gated and free-flow parking systems. The Einar camera ensures precise license plate recognition at city speeds and real-time speed measurement through video analytics. Its energy-efficient design supports solar power, making it versatile for various parking environments. For parking enforcement, the compact Lynet camera offers a highly accurate mobile ANPR solution, easily installed in patrol cars for efficient monitoring and compliance.

With a blend of local expertise and global reach, Adaptive Recognition deliver seamless, efficient solutions backed by personalized support.  

Contact us to discover how we can optimize your parking systems!

About Parking Network 

Parking Network is the leading information source for parking. We connect professionals through our online and offline channels. On our online platform parking.net, we publish the latest news, product showcases, job openings, and tenders related to parking. On our platform, you can access the most complete industry and parking professionals directory. Offline, we host niche-specific networking events such as the Airport Parking Network Event.