Real-virtual fusion model for traffic animation

In this paper, we present an innovative, animated traffic simulation method that we designed to feature an enhanced sense of reality and diversity of traffic flows. Instead of the typical one-off initialization, our simulation method includes continuous, real trajectory data input providing an interactive control function that maximizes the characteristics of real-world traffic flows. Our fusion models represent a comprehensive integration of the interactions among real-data-driven and virtual vehicles, thus depicting accurately the irregularity of traffic flows. Test results showed that animations generated via our proposed method depict inverse and irregular vehicle driving behaviors throughout the entire traffic flow.