Influencing Factor Analysis for Visual Information Search While Driving: Based on Modeling of Visually Interesting Regions

Many methods and eye movement indexes were used to describe drivers' visual search. But drivers' visual search behavior should be a four-dimensional index, which not only includes fixation distribution, but also includes fixation duration.Thus, it is difficult to find the regularity of drivers' visual search behavior by single factor analysis for drivers' eye movement index. In order to quantitatively analyze the characteristic of drivers' visual search behavior, the paper introduces a new index, called the Visually Interesting Region (VIR). VIR includes area and barycentric coordinates two parameters, and is a new index of visual information search while driving. According to the analysis, the paper found area of VIR and barycentric coordinates separately has correlation with horizontal alignment and traffic flow, of VIR mainly affected by highway alignment and traffic flow.