Machine Vision to Alert Roadside Personnel of Night Traffic Threats
Liang Wang, Berthold K. P. Horn · IEEE Transactions on Intelligent Transportation Systems · 2017
In the United States, every year, several people whose job takes them to the sides of roads, are injured or killed by roadside collisions. This could be avoided if a warning signal could be sent to them. In this paper, we describe a machine-vision based alerting system which detects and tracks headlamps of cars in night traffic. The system automatically computes a “normal traffic” region in the image. Unusual trajectories of cars are detected when the images of their headlamps move out of that region. The system promptly sends a warning signal once a risk has been identified. The system runs on the Android smart phones, which are mounted on cars or on roadside fixtures.