AI Based Object Classification

The classification of objects, in conjunction with sensor fusion, are hot topics for most automotive sensor suppliers nowadays. The final goal is fully automated driving. With the help of artificial intelligence (AI), implemented in (deep) neural networks (DNN), so far unknown performance levels of object classification can be achieved.

In smart cities the optimization of traffic flow is one of the most important aspects. For that, smart infrastructure sensors are required which can measure the traffic flow. smartmicro sensors can do this very precisely and provide real-time object data which are fed into local intersection controllers for direct local adaptive traffic control. The traffic data is transmitted to a traffic operation centre (TOC) and allows an optimum traffic flow in a certain region or quarter or on a longer arterial, where the traffic flow is remotely (but automatically) controlled from that TOC. Real-time object data can now also be directly transmitted over a Car2X wireless transmission from an intersection to connected vehicles. This avoids accidents and protects vulnerable road users (VRUs) such as pedestrians and bicycle riders.