For today’s systems integrator (SI), the ability to offer a complete security portfolio can be a key differentiator. While SIs have been able to support video management and access control for years, adding automatic license plate recognition (ALPR) solutions was generally considered more difficult. This is because, traditionally, ALPR sales were a specialized area that required more technical training and because the cost per lane seemed prohibitive to many end users.
Yet, advancements in ALPR mean that these barriers are no longer in place. Improvements in the technology have brought down the cost per lane, making ALPR solutions more attractive for use in everyday applications. And increased system flexibility, as well as reduced installation and deployment requirements, mean that this is a good time for SIs to add ALPR to their list of offerings.
The first step to getting started is understanding the basics. Let’s get started.
What are server-based systems?
When it comes to ALPR, not all systems are created equal. Where server-based solutions may seem like an easy option, this is not always the case. Server-based ALPR involves using traditional security cameras to capture and send images to a server where the plates are read. This requires purchasing and installing ALPR software on the server to allow it to find a good image of the license plate within the stream and then make a determination about that plate.
Using traditional security cameras might appear to be a cost-effective choice, particularly since many end users already have them installed to monitor the side of a road or driveway. But these surveillance cameras are typically optimized for an overview. To read a license plate, a camera must be able to view the front or back end of any vehicle.
While it is possible from a technology perspective to use a traditional camera, the end user has to readjust it down to where it can capture a plate. This often defeats the purpose of using existing security cameras because, once you focus it down to get the pixel count necessary to accurately read license plates, the camera no longer provides the overview it was intended to capture.
Most surveillance cameras are not designed to capture details as specific as numbers and letters on a license plate. They typically use low-cost rolling shutters instead of higher performance global ones. This can result in reduced read accuracy at night, at high speeds and under a variety of weather conditions.
One solution to this problem is to install an additional surveillance camera to ensure that the area is covered properly. However, instead of saving money by using one camera for two purposes, costs are increased because you have to purchase the server and software license necessary to read the license plate as well as any additional cameras that are not, ultimately, purpose-built for the task.
When data can get lost in transmission
Another problem with server-based ALPR is that the raw footage from the camera has to be transmitted to the server for processing. The recurring costs associated with transferring all the data necessary for the server to get a good read of a plate can be high, especially if it is being sent over a cellular network. Given the fact that surveillance cameras transmit a constant video stream even when no vehicles are present, it becomes clear how expensive server-based ALPR can be.
This process of moving data is also prone to transmission problems. If communication between the camera and server is interrupted or dropped, the footage disappears and can’t be used for real-time identification. It also means that the data won’t be available for any post-event investigation because, when the transmission is interrupted for any reason, the video is lost as nothing is stored on the camera.
The advantages of edge-based systems
The main difference with edge-based license plate recognition is that it uses ALPR cameras purpose-built to read license plates. With this solution, the algorithm that processes the raw video stream and reads the license plate is in the camera itself.
One key advantage of using purpose-built technology instead of adapting video surveillance cameras is the increased accuracy in license plate reads. With an edge-based ALPR, end users get a system that is designed to address their specific license plate recognition needs.
What Integrators Should Look for in an ALPR Vendor
When choosing a manufacturer that can support your needs, you should be looking for a vendor that produces the hardware as well as the algorithms and firmware that go along with the cameras. This ensures that the software will work hand-in-hand with the hardware’s engineering, which will ultimately ensure the best performance from the technology.
You should also look for a manufacturer who has been committed to the development of ALPR technology and does not just regard it as a side business. And they should have the resources to support you on your initial sales presentations and deployments to make sure that you are successful with your initial rollout as you ramp-up in your ALPR business. Now is the time to add ALPR to your security portfolio.
In many jurisdictions, the variety of plate designs can make it difficult for traditional security cameras to get an accurate image. Plates can be flat or embossed and can have different reflective values. All of which can impact the quality of the image capture. And, if the camera is not able to capture a clear image of the plate, then the processor will not be able to make an accurate read.
Shutter speed plays an important role in capturing good quality images. With traditional image-making, the shutter speed controls how long the lens is open to incoming light. When the shutter is open longer, more light gets in. The same is true with ALPR. If the shutter isn’t open long enough, images can be too dark to read; if it’s open too long, images can be flooded with light or blurry because of vehicular motion, which also makes them difficult to read.
In a purpose-built ALPR camera, the shutter speed is constantly adjusted to compensate for different lighting conditions. For example, an ALPR camera taking 30 images a second will adjust its shutter speed for every one of those images. This is because the goal is not to take the clearest picture of the surroundings but to capture the most accurate image of a license plate.
The camera works in conjunction with the firmware to adjust the illumination and shutter speed to get the best image for an accurate read. After taking several pictures with different shutter speeds, the camera goes through the images to find the one with the highest quality and then gives that image to the camera processer to read the license plate.
With edge-based ALPR, all of the image processing is done in the camera at the edge. This means that the camera transmits a drastically reduced amount of data over the network, namely the license plate read, a compressed snapshot of the vehicle, and associated metadata, including date, time, and location. This requires much less bandwidth and expense than transmitting live video feeds.
In addition, because the processing is done in the device, even if the network goes down, the camera continues to read images and collect data. Then, it can send all the stored information when communication has been re-established.
How machine learning can help
Machine learning is making ALPR more attractive to end users and easier for SIs to integrate into their portfolios. In the past, image processing was done through rules-based algorithms. Data scientists wrote the code that enabled the computer to determine if a character was an X or a Y.
Sometimes, a change in the design of a plate’s background could interfere with the read. In the state of Oregon, for example, they developed the Crater Lake plate with a mountain range in the background. Using rules-based algorithms, some ALPR systems read the I as a T. And this is not an isolated problem.
Many states in the U.S. have 80 to 90 different license plate designs. You can have one that indicates your alma mater, your profession or your favorite football team. Features on these plates can significantly affect readings, and as a result, data scientists have had to write more code for the rules-based algorithms to account for all the changes in pixels around the letters and numbers on each plate.
Now, with machine learning, data scientists simply write the code that allows a computer to learn to read plates. During supervised machine learning, the camera captures images, and then a trainer labels these images, for example, indicating that this is a letter I.
When the camera has ingested enough labeled images, the algorithm allows the computer to look at an image and make a mathematical guess as to what the letter is. In the case of the Crater Lake license plate, the computer could be trained to recognize that the mountain range in the background is not part of the letter I.
By moving to machine learning, ALPR has gone beyond what rules-based algorithms could identify. Today, ALPR cameras can read other features from video images, like the color of a car and its make. This means that, if a witness to an accident cannot remember a car’s license plate but does know that it was a pick-up truck or red Toyota, investigators can use these descriptions as search criteria and the system will be able to pull up each instance from its feeds.
Who is using ALPR technology?
Law enforcement and parking were the first verticals to adopt APLR cameras and systems. The police started using APLR technology in patrol cars to read license plates and compare them with a hotlist of wanted or stolen cars. Parking enforcement then adopted the technology to help manage permits and identify parking violations.
But ALPR is now being used in other verticals. This is, in part, because machine learning has increased the read accuracy of purpose-built cameras to the same level as other technologies, including RFID. Government organizations are now expanding their use of ALPR to include different applications, like tolling. In addition, the private industry is starting to adopt the technology for a variety of uses, including customer convenience, loyalty rewards programs, and data research.
There are two other reasons why ALPR sales opportunities are increasing. The first is that it is easier than ever for an SI to install an ALPR solution, which also makes it easier to talk to prospective customers about different possible applications. The other reason is that the cost per lane has dramatically decreased over the past few years.
This has opened the technology up to more mainstream applications, whether it is a supermarket that reads customer license plates and then delivers pre-ordered groceries directly to their car or a car wash that administers a membership reward program using ALPR cameras.
How does an SI add ALPR?
SIs are now finding it easier to distinguish themselves from other integrators by including ALPR as part of their offerings. This is great news for end users. When an SI can offer customers an end-to-end solution, users can extend their security beyond the exterior walls of their buildings. Instead of knowing who is on-site only as they come through the front door, organizations can now use ALPR to understand who has entered their property. This pushes the boundaries of their physical security system to the edge.
If you are an SI who is considering adding ALPR to your security portfolio, the first thing you should do is look at the options. Compare server-based and edge-based systems to understand the benefits of these technologies for the verticals that you serve. Once you have decided on the type of solution that best suits your specific market, consider standardizing on a product or manufacturer.
Working with a manufacturer that can help during presentations, and especially during your first six months of taking on an ALPR product line, can be beneficial. You need a manufacturer that can support you through education, particularly around product nuances. This can help you correctly set customer expectations at the beginning of a project.
It’s important to remember that, with ALPR, customers know what they want. They know the data they want to extract from their systems and the performance they require. It is easy to see if a set-up is working properly because the system itself produces validating statistics. If a hundred cars go by an ALPR camera and the camera reads only 30, the customer can see this. Setting clear and manageable expectations will lead to greater success for the customer and profitability for you.
Larry Legere is the Commercial Director of AutoVu at Genetec.
The post Are You Ready to Add ALPR to Your Security Portfolio? appeared first on Security Sales & Integration.