Car Radar

We assist in tracking and finding stolen vehicles.

Our mission is to increase the rate of success during search for a stolen car, bike or bicycle. We want to achieve that by encouraging car drivers around the world to participate in our smart network that uses advanced machine learning techniques to recognize license plates and match them against our database of stolen vehicles which is maintained by our end users. Drivers can install our hardware into their cars and assist in finding stolen cars. We make use of Raspberry PI for this. We provide attractive bonuses for lucky drivers which were able to spot license plates from our list of stolen vehicles. We can provide sticky numbers for bicycles as well, since they typically do not have recognizable numbers.

Used technologies

license plates detection technology: https://github.com/openalpr/openalpr

Our system consists of three main components:

  1. Raspberry Pi client for processing raw footage from a dash cam.
  2. Web and Android clients for end users that helps you search for stolen vehicles
  3. Backend for processing requests and database maintainance

Raspberry Pi client

Allows to use internal/external webcam api to get videostream for future processing. For image processing we use modern technique, based on neural networks image recognition. License plates found on processed images are then transferred to backend server.

This should make use of street and road cameras, which are already used for different purposes in many cities around the world and are becoming even more wide-spread.

Web and Android clients

Client application for people, who got their cars stolen. We understand, that it is really important to do anything possible in a shortest period of time after theft, so our apps is user-friendly and extremely simple to use. You just need to enter your license plates and personal document data (for security), and we immediately start searching for your car. When it is found, you will be notified in an app and via email.

In our world mobile applications are taking more and more of people's lives. So we believe mobile application is the best way to provide service like ours.

Backend

This part of our system is responsible for data storage and notification of our beloved end users! We keep all the observed license numbers in order to be able to handle the future query. We maintain the license number along with the last observed location (longitude, latitude) on the map. Therefore the user can get a clue on the possible location of the stolen vehicle by seeing the latest position. The user can subscribe for the future matches as well in case our system still has no information about it so he never misses his stolen vehicle.

Future work

  1. Security support. It is very important part of our service, as nobody wants anyone else to see their property on the map. With access to official database of vehicles our service will become 100% secure and reliable.
  2. Nowadays many drivers have DVR in their cars. Many of these devices work on Android. We plan to release Android API which should allow to use DVRs as video streaming devices. That would let us get thousands more relevant images of vehicles. Imagine, how higher will become the chance to find any stolen one!
  3. Car model, color, etc detection: this will improve finding accuracy and help people recognize their cars on images.
  4. Drivers loyalty program. We believe, that every good deed should be thanked. And we are really thankful to all the people, who let us use their devices to help others find thieves, so we want give them something back.
Share this project:
×

Updates