Risk assessment of Highway Safety Based on Artificial Intelligence: A New Method of Highway Risk Assessment

Abstract

A new risk assessment method is proposed which combines deep learning, computer vision and traffic safety. The method consists of several models. The highway risk factors identification model, extracts risk elements from video images. The vehicle velocity speculation model is built for the driving speeds calculation of different vehicle types. The Road alignment computational model is built for calculating highway alignment according to GPS. The highway risk management model for evaluating highway risk grades and factors is given. The new method greatly accelerates the speed of highway risk assessment and reduces the human cost. For the overall risk assessment, its accuracy depends on the results of each module. Through experiments and engineering applications, the new method can achieve the same or even better effect as the expert manpower.