Could a “Monocular Advantage Effect” Be Measured in Driving Simulation?

Abstract

Illusory depth perception experienced in driving simulators is afforded by monocular depth information contained in visual displays. Presumably binocular convergence and binocular disparity, though useful for depth perception in real environments, may poorly contribute to illusory depth in a driving simulator. Instead, they may generate conflicting information by revealing the distance of the display screen and its flatness. Nevertheless, illusory depth induced by monocular information contained in visual displays usually produces enough immersion and realism to create the illusion of driving in a real environment.

Many authors have noted improved depth perception in paintings, photographs, and even in drawings when viewed monocularly. However, this effect, known as monocular advantage, has never been explored in driving simulation. The purpose of this experiment was to assess whether the effect might exist in driving simulation. It was expected that drivers would perceive distances in depth better and more accurately with a monocular than with a binocular viewing of the display. Distance estimates were evaluated for two types of driving maneuvers referred to as alignment and bisection. Results showed that when significant performance differences between monocular and binocular viewing conditions occurred, target cars were perceived farther in depth and more accurately using monocular vision.

Alternative viewing conditions using both eyes are discussed at the end of the article.

Notes

¹An objection could be raised against our procedure consisting in covering one eye to produce monocular vision. It could be argued that one eye viewing does not cancel completely binocular information for viewing distance. In fact, vergence, which is cross-linked with eye accommodation, is not deleted with this procedure. Then, because this binocular cue for distance remains in one eye viewing we should have hypothesized that there is no difference between depth viewing with one eye and that with two eyes. However, our results reject such hypothesis and reveal that perceived distance is better with one-eye vision than with two-eye vision. We should then conclude that that vergence linked with accommodation was not strong enough in one-eye viewing to cancel monocular advantage effect.

²Using binocular data only, we published a paper on distance perception in which we discuss in detail the effects of speed and type of task on distance perception .

³Data obtained in this first phase are not directly comparable to the others. In fact, the first target vehicle (car B) was located at apparent distances of 240 or 260 m from the starting point, as compared to 40 or 60 m from car B for bisection and the same distance from C for alignment. Because of these distances, the driver tended to approach the target position much more rapidly than in other situations. Furthermore, contrary to the other conditions, he had no ground “signal” available to warn him of the relative proximity of the target. For these reasons, we have decided to ignore results from the first phase; nevertheless, it allows us to apply uniform approach conditions (speed and distance) to the other tasks (bisection and alignment).