How LPR works – The best tutorial for License Plate Recognition

Before decide for the best vehicle license plate recognition system (LPR) it's important to learn what factors you must take in consideration.

Be aware that LPR depends on good Optical Character Recognition (OCR) systems, so choosing good OCR services to work with is essential. 

There is no straightforward answer on which is the best camera or system for vehicle license plates recognition, it depends on your project's budget and what you really need and expect from your system.

See in this picture the example of SECUROS software used for license plate recognition. To learn more, visit the official ISS website at: http://isscctv.com

There are two ways to do this job of capturing and recognizing vehicle license plates, and you must decide which one is most appropriate for your case based on your budget and type of implementation.

It is possible to purchase all items separately and work on  a solution integration by installing the camera, software and accessories or buy a LPR camera that already brings everything integrated into a package.

In this article, I will talk about the first system. It is possible in this case to merge different technologies, IP cameras brands and LPR software to better meet the project's specification.

If you choose this option to have your license plate recognition working correctly you need to understand what features a camera must have to allow the correct settings according to the project.

Manual integration requires cameras, protection boxes, switches, cables, installation accessories, as well as servers, software and specialized workers.

It's necessary to install all the devices and adjust the IP camera and LPR software parameters. It's a really time consuming task

The diagram below shows an example of a traditional license plate recognition (LPR) system with its main components.

In this model, the IP cameras must capture the image with enough resolution and quality so that the LPR server can use the optical character recognition (OCR) and store the information in a database. A monitoring station can display the images and the license plate recognition results.

Another great advantage in the use of IP cameras is the possibility of using embedded intelligence, there are some LPR software developers that have the option for two steps processing by using the IP camera to process part of the job and let the second part to be processed by the LPR server.

To learn more read the article: Video Analytics for CCTV cameras.

Another important factor to consider in the switches for your project is Power over Ethernet (PoE). To learn more read : PoE Switch for IP cameras.

To obtain an image with enough quality for the LPR software do a good job, the camera must have a minimum image resolution.

Even analog cameras with traditional resolutions can be used for LPR, however it is recommended to use IP cameras with high resolutions.

==> Watch the video below to learn more about lenses for CCTV

==> Watch the video below with an example on how to adjust the shutter speed on the camera and fix the issue with blurry images.

Depending on the scenario it's possible to use different types of features according to the incidence of light, shadow, excessive light or reflections.

If you want to learn more about light compensation, watch this video that explains about BLC, WDR and Sens-UP. Keep in mind though that the Sens-UP is not very useful in low-light and motion scenarios, which may be the case with vehicles on a dark road. In scenarios like these it is important to use additional lighting, whether normal or infrared.

It is important to know that there is no specific camera for vehicle license plate recognition that is best for all situations, it will all depend on what is required for your specific project, you can choose different types of cameras and lenses that will work just fine as long as you understand the concepts.

With this software you can work with information such as distances from the camera to the vehicle, lens focal length, camera resolution, and more.

Several factors will influence how well your system captures and recognizes the license plates, you must be aware of each one of them and make the necessary adjustments to optimize the system, here are some important details from in the ISS software manual .

When installing the camera in an upright position as shown in the photo hold the maximum angle at 40 degrees so that the plate recognition software can operate efficiently, smaller angle is even better.

When installing the camera in a side position as shown in the picture hold the maximum angle at 30º. Smaller angle is even better.

In summary to have a LPR working properly, follow the recommendations:

In another article I will talk about professional cameras with embedded LPR technology such as Hikvision, Vivotek and Genetec cameras. This type of camera facilitates the installation because it is already prepared with the technologies that were discussed here in this article