Visual Information in Judgments of Head-On Collisions

Abstract

Head-on collisions result in a substantial number of fatalities. To detect head-on collisions, drivers must judge effectively the direction or heading of their own vehicle in relation to the heading of oncoming vehicles. In our previous study, we used computer simulations of self-motion through a traffic scene to measure judgments about whether a head-on collision was imminent. Results suggested that judgments about head-on collision are affected by both the optical flow information provided by the centerline and the optical flow information provided by the oncoming car. The objective of the current study was to further examine the effect of different components of the optical flow pattern on judgments of head-on collisions. We measured judgments about head on collisions while manipulating local optical flow from the oncoming car and global optical flow from the background scenery. Our results suggest that visual information about the oncoming car's motion was more effective than visual information about self motion. The implication is that it may be beneficial for drivers to focus greater attention on the information about the oncoming car's motion in order to improve judgments about head-on collisions. Further research is needed to evaluate this possibility.