License plate recognition (LPR) technology has become a staple in law enforcement. The powerful addition accelerates investigations and lead generation. While, the vast market of LPR systems operate off of similar setups, today, we want to talk about the evolution of LPR software and how agencies can take advantage of those technological changes.
This blog is going to cover:
- LPR technology
- OCR vs. AI software
- Use Case: Using AI for LPR with Rekor and Rekor Watchman
License plate reader technology captures license plates and, depending on the system, vehicle information (make, model, color). To capture the vehicle, systems utilize a combination of mobile or fixed cameras, a processing server and backend software.
The data gathered from provides extensive benefits including (but not limited to) analytics, area specific movement tracking and hot list vehicle notification. Overall, LPR technology aides in increasing lead generation, leading to more solved cases.
OCR vs. AI
A key differentiator in platforms is the backend software used for capturing and identifying plates. For more than 20 years, Optical Character Recognition (OCR) technology has been the prominent. OCR recognizes numbers and letters through images from high-end, expensive hardware with the accuracy range average between 80-85%. In addition to the high cost, the hardware is also limiting the data gathered due to the camera having to be focused on a specific location. While OCR has provided valuable technology to the LPR industry, it has limitations compared to new software.
Artificial Intelligence (AI) and machine learning brings an innovative approach to license plate recognition. A simple way to understand how AI works with LPR is – if the human eye can see it, the AI software can read and capture the characters on a plate. AI offers the capability to use both images and video for reads leading to a 99.02% accuracy rate.
A big differentiator on the two options is the cost. In the public safety space, we understand that departments are on a fixed, allocated budget therefore maximizing a product within a cost is critical. OCR solutions require expensive, single purpose hardware, which then leads to limited deployments. Whereas AI solutions can utilize existing commodity hardware.
Use Case: Using AI for LPR with Rekor and Rekor Watchman
In case you missed the news, we recently partnered with a new LPR provider – Rekor Recognition Systems. The Rekor Watchman solution implements Open ALPR software using AI and machine learning to capture license plates.
The Rekor Watchman software provides valuable insight beyond a simple plate read and image. Other vehicles variables are captured natively, just like the human eye processes an image. Rather than running a license plate through the system to match a VIN number, the software reports on what it natively sees in the image or video. Variables collected include vehicle, direction of travel, time of read and confidence of the read.
Rekor’s innovative approach is changing the LPR space. The enhanced capabilities and accuracy from AI and machine learning in combination with the opportunity to utilize existing IP, traffic or surveillance cameras helps to maximize the investment for departments.