There are two parts to LPR: the act of reading the plates and the act of comparing the plate reads to a database. The latter distinguishes different applications including scofflaw and wanted vehicle identification, parking permit enforcement, vehicle inventory and access control. The reading portion actually involves three sequential actions: identifying a licence plate in an image by looking for rectangles; once identified, segmenting the alphanumeric characters; and lastly, using optical character recognition (OCR) to read the characters on the plate. This complex process is optimised with specialised infrared LPR cameras and backend software for the management of the plate reads.

Ideally the LPR software is integrated with a video management system (VMS) or can be added as a module to the client’s existing VMS. This will allow the LPR system to index reads with correlated video footage for further verification and investigation. Now, what kind of applications can LPR be used for? Even though LPR cameras can be mounted on vehicles for mobile applications, for the sake of the article, I will only focus on fixed applications.

The most common applications are “hot and cold” lists for vehicles of interest. For example, a maritime port authority may have a list of vehicles they suspect of engaging in criminal activity. Therefore, they can set up a fixed LPR system at the entrance of their property. If a vehicle of interest nears or enters the premises, an alarm will be triggered in their CCTV control station, showing a colour image of the identified vehicle, and its licence plate, and even prompting the actual video footage of the driver and the car.