In this paper, the author demonstrates that the segmentation problems related to object-motion detection increase with simultaneous subject-motion, leading to higher detection times (experiment 1). Motion detection is impaired mainly by the optical flow, associated with self-motion, surrounding the leading vehicle. The general optical flow in the visual field and the peripheral one seems to be less important (experiment 2). It should be noted that the available hardware clearly constrains the density of optical flow that can be stimulated. Therefore, it is reasonable to expect higher rates of detection times under natural conditions and with more powerful experimental facilities. Considering the results in these two experiments related above, the author outlines two applied implications. First, the current models of computation of braking and safety distances should be reviewed. The driver's reaction time should be considered as a variable and not as a constant. The degrees of freedom would be the subject-motion velocity, the relative motion of the target vehicle or object, and the density of the optical flow surrounding the target. Secondly, it is reasonable to expect that by manipulating the road environment the driver's motion detection performance could be controlled. It could be possible to promote better driver reaction times by reducing the road texture and the amount of traffic signs, both horizontal and vertical. However, further experiments should be performed to enable the development of a new model of driver reaction time computation and specific guidelines to road design.
Samenvatting