AI-TRAFFIC, the video analytics solution from A.I. Tech for traffic monitoring, now integrates a new vehicle classification mechanism based on the use of deep neural networks.
The solution recently awarded with the IoMOBILITY AWARD in the Mobility Analytics category, allows you to detect, counting, classification vehicles (therefore distinguish cars, motorcycles and trucks). It also allows you to identify the average speed and colour for each vehicle.
AI-TRAFFIC also allows the detection of potentially dangerous situations on roads, such as the presence of a pedestrian on the road or a vehicle stationary in a prohibited area, congestion, vehicles moving at high a speed or finally vehicles in a wrong direction.
Using deep neural networks substantially improves system performance during the classification phase, and at the same time is useful to reduce the number of false alarms generated. Another huge advantage lies in the processing times: in fact, thanks to the use of optimized proprietary networks of A.I. Tech, the system allows classification also if the application is working in edge mode, directly on board of the camera.
