Method and apparatus to prevent VRU accidents

An apparatus configured to prevent at least one vulnerable road user (VRU) from having an accident, the apparatus comprising:

• a driver monitoring unit configured to utilize at least one sensor provided in a vehicle to detect a gaze area corresponding to a gaze direction of a driver of the vehicle and a head pose corresponding to a direction in which the driver's face is directed by calculating various distance differences based on a plurality of landmark points detected from the driver's face image;
• a VRU monitoring unit configured to calculate position information including current position and predicted motion information of the VRU existing in a peripheral area of the vehicle, the VRU is detected by using the at least one sensor;
• a detection unit configured to detect whether the driver is aware of the VRU and whether the VRU is aware of the driver based on a gaze area of the driver, the position information and heading information of the VRU, wherein the heading information is predicted by a future position of the VRU and an expected time to reach the future position;
• a warning processing unit configured to output warning information on a dangerous situation to some or all of the driver and the VRU in response to a detection result of the detection unit;
• a vehicle monitoring unit configured to predict a future movement of the vehicle based on a vehicle movement and a driver control input detected by using the sensor;
• and
• a collision risk prediction unit configured to determine whether a collision risk is present between the vehicle and the VRU based on the future position calculated by using the VRU monitoring unit and the future movement of the vehicle predicted by using the vehicle monitoring unit.