Visual Search and the Detection of Collision Events during Driving

Abstract

In the present study we examined the accuracy of observers in detecting collision events. Observers were presented with displays simulating a 3D environment with multiple moving objects. We used a signal detection paradigm to determine the ability of observers to detect a collision event when multiple moving objects were present. We also examined whether the task can be performed in parallel or requires attention. The results showed that observers' sensitivity decreased with an increase in the number of distracter objects moving in the scene, suggesting that observers were performing a visual search to determine whether a collision target was present in the display. This finding indicates that collision detection when multiple-moving objects are present is not processed in parallel. In addition, the results indicated that sensitivity increased with an increase in display duration. This suggests that as the collision event became more imminent detection performance increased.