We are developing a system to track and extract parameters of vehicles via traffic cameras. The main parameters are license plate, car brand, car color, and moving direction. Base on this information, the system can analyze and detect violations such as red light violation, wrong lane or illegal turn.
The tracking system can now detect and track up to 30 vehicles in real-time on T4 GPUs.
The recognition system can detect 8 most common car brands: Toyota, Chevrolet, Ford, Honda, Hyundai, Kia, Mitsubishi, Mazda and 8 most common colors: black, blue, brown, green, orange, red, white, yellow.
This system is a combination of many advanced image processing techniques and state-of-the-art deep learning models, focus on object detection, object tracking and text recognition technologies.