Vehicular collision avoidance using video processing and vehicle-to-infrastructure communication

Road intersections, without signal lights, especially those with walls or other objects that block the driver´s view, are a major traffic safety hazard. A significant number of accidents happen at these intersections. The risk factor is further aggravated by the presence of animals or pedestrians which is a common occurrence in developing countries such as India. In our proposed research, we aim to provide early warning to drivers at intersections to help them avoid possible collisions. This can be achieved with the help of a low-cost camera unit combined with special hardware for image capture/processing and Vehicle-to-Infrastructure (V2I) communication. Therefore, when an object on a collision path, such as a motor vehicle, pedestrian, or animal is detected by the system, a warning is issued to the nearby vehicles enabling drivers to react to critical situations quickly and effectively.