CarGate

You are here: Home > Projects > CarGate

Cars have a number of unique characteristics that can be used for identification. License plate recognition is the most commonly used method to identify a car. Since this method is camera based, difficult weather and light conditions have a negative impact on the recognition rate. This project is about using different characteristics such as the sound pattern, magnetic spectrum, and even the color of a car to improve the recognition rate.

Sensor Fusion

CarGate People are excellent at recognizing the sound of a car from other sounds. They can even distinguish roughly the type of car that is passing by. INCAS³ aims at developing a sensor system that can perform a similar task; identify a (specific) car based on the sound it makes.

The metallic structure of a car creates another possibility for identification using 3D magnetic field sensors. An advanced magnetic field detection system can differentiate between large iron parts, such as the motor block and axles. Smart software then assesses the data to see whether it is a passing car, a truck, or a motorcycle. The software may even become smart enough to detect vehicle brands and models.

By combining the data from the magnetic field sensors with the data from the sound sensors, it becomes possible to map out traffic streams at an intersection, for example.

Combining the information obtained from the sound and magnetic identification with license plate recognition, it becomes possible to achieve almost 100% vehicle recognition.

Partner

Within CarGate, frontier scientific knowledge from the INCAS³ Cognitive Sensor group is combined with the proven industrial expertise from Parkingware. INCAS³ will focus on the development of novel characterization methodologies for sound patterns and magnetic spectrum recognition. Parkingware focuses on improving their license plate recognition system.