VEHICLE RECOGNITION SDK

Integrate ALPR and vehicle recognition into your apps

Our Vehicle Recognition SDK is a C/C++ library, with additional bindings in C#, Java, and Python, that reads from images or real-time video streams, processes the data, and produces license plate and vehicle recognition results.

The Vehicle Recognition SDK runs locally on Windows, Linux x86, and ARM while integrating directly with your application.

Learn more

Request SDK license 

Two cars driving on a road with their license plates being recognized and displayed in green info boxes

CUSTOMIZABLE RECOGNITION

Develop on-prem using two unique recognition APIs

While running locally on your hardware, the Vehicle Recognition SDK exposes two unique APIs for identifying license plates and vehicles from either still images or real-time streaming video.

The Vehicle Recognition Image API recognizes license plates and vehicles through individual images, while the Vehicle Recognition Stream API is used for license plate and vehicle recognition on videos or time-ordered image sequences.

Read documentation 

Packed with Features

Direct access to our recognition libraries



Operate on-premises

No need to send data to our cloud, as our Vehicle Recognition SDK runs completely local on your own hardware to securely integrate directly into your applications.



Code in multiple languages

While our SDK is designed to work best with C/C++, additional bindings are available for C#, Java, and Python, allowing you to develop in the language you are most familiar with.



Utilize varied sources

Whether you are sending still images or real-time video streams of vehicles, our Vehicle Recognition SDK has the ability to process both using two unique APIs and frameworks.

Supported Countries

Recognize varying plate styles across the globe

Our Vehicle Recognition SDK contains specialized training data for many license plate and number plate styles. If your country is not listed, you may still experience high accuracy rates by selecting a country with similar looking plates.

Primary regions supported:
USA, Europe, Canada, Brazil, Japan, China, Australia, Mexico, UAE, Indonesia, Saudi Arabia, India, South Korea, Thailand, Argentina, New Zealand, Great Britain, South Africa, and more

View complete list

Request SDK license 

License plates from around the world mounted to a wall

Blue Dodge Challenger parked on the side of the road with a green bounding box around it

./alpr_c_sample  cameraimage.jpg
plate: G33FGW - us-nj (98.9336195678711)
  - make: dodge (89.90132690429688)
  - color: blue (96.43530151367188)
  - orientation: 0 (96.70639038085938)
  - make_model: challenger (91.49954223632813)
  - body_type: sedan-standard (96.80194854736328)
  - year: 2010-2014 (73.43781455078125)

OUTPUT EXAMPLE

Easy-to-read and leverage recognition results

This example shows a standard response from processing a still image of a vehicle with the Vehicle Recognition Image API within our complete Vehicle Recognition SDK library.

Both APIs included in our Vehicle Recognition SDK can provide license plate number and state of issuance, as well as vehicle make, model, color, body type, year, and orientation.

View code samples 

Vehicle Recognition SDk

Frequently asked questions

Have a question that isn't answered below? Be sure to check out our Support Center.

Can I use the SDK without an internet connection?



Yes. After installation, the software does not require any internet connectivity. The license plate and vehicle recognition data may be captured and made available to you entirely offline.

Can I embed the library into my proprietary application?



Absolutely. Rekor's recognition software can integrate with other applications in a number of ways. Our Rekor Scout^® Local Agent can be configured to read data from a video stream and send all license plate data to another application via HTTP POSTs. The Vehicle Recognition SDK is a code library that provides direct access to our APIs to analyze still images or real-time video streams.

Our code library is directly compatible with C/C++, C#, Java, Node.js, Go, and Python code.

What countries do you support?



We support nearly 70 countries and over 500 regions all over the world with characters in many different languages. If your country is not specifically supported, it may work well with another country that has a similar design.

What cameras work best for real-time video stream analysis?



We can connect to any IP camera that supports H.264 and MJPEG streams.

What if I have technical or installation questions?



We are always here to help and most answers can be found on our Knowledge Base. If you can’t find the answers you are looking for, please Contact Us directly.

Get In Touch

Ready to integrate vehicle recognition into your apps?

Our sales representatives are standing by to help craft a plan to fit the needs of your business. We currently have clients processing 100+ million recognitions per year and are adept at deploying enterprise-level volume.

Man sitting at a desktop computer with two monitors looking at a dashboard on one monitor and a vehicle render on the other

Want the future of roadway intelligence delivered directly to your inbox?



Rekor Systems Reports Third Quarter 2024 Financial Results