Design on Freeway Automatic Incident Detection System Based on Edge Computing and AI Algorithms

Abstract

Accurate, fast and real-time freeway automatic incident detection can improve the freeway intelligence level to a great extent. Based on edge computing and artificial intelligence algorithms, the framework of freeway automatic incident detection system is proposed in this study. The core structure relying on edge computing is presented initially and a series of intelligent algorithms are employed to extract, identify, track and predict the trajectory of the object vehicles. Various scenarios of incidents are detected and pre-warned. The testing results reveal that the system can provide higher accuracy for the incident detection purpose.