Multi-sensor environmental perception and information fusion for vehicle safety

Abstract

In the progress of improving the safety of a vehicle while driving, multi-sensor environmental perception has many advantages compared with other approaches. This paper is devoted to proving that multi-sensor information fusion that is used on a vehicle can help identify and eliminate the spurious data, which is obtained from multi-sensors, based on Bayesian approach. And it is one of the most important but simple approaches in the information fusion field. The results of the simulation show that centralized sensor fusion scheme using Bayesian approach has higher efficiency compared with other methods.