A System for Risk Identification of Pedestrian-vehicular Collisions and Intelligent Control

To avoid pedestrian-vehicular collision,systems for pedestrian detection and collision warning are used in intelligent driver assistance systems.However,the warning systems can be useless under impaired driving.In this study,based on a pedestrian detection system,an intelligent control system for vehicles is proposed to avoid pedestrian-vehicular collisions.First,Histogram of Oriented Gradient (HOG) and Support Vector Machine (SVM) are applied to detect pedestrians.Risk of pedestrian-vehicular collisions is predicted by using an algorithm of risk identification.Then,based on a vehicle dynamics model,speed of vehicles is adaptively controlled under different risk levels.The results show that the developed system can detect pedestrians quickly and accurately,its accuracy can reach 89% in real-time under dynamic scenes.The risk warning system can control the vehicle to take a safe and stable brake under emergencies,thus pedestrian-vehicular collisions can be effectively avoided.