Abstract
Intermittent occlusion has little effect on the detection of a target moving in a fixed direction embedded in random motion noise (Watamaniuk & McKee, 1995) but the perceived speed of such a target may be consistently underestimated (Hecht et al., 2002). The present study sought to determine whether this perceptual speed reduction is due to the occlusion per se or to poor local motion signals.
In each trial, an observer saw a moving target spot that was intermittently occluded and then a moving target spot that was constantly visible. Target direction in the first interval was randomly selected to be leftward or rightward while that in the second interval was the opposite. The constantly visible target moved at 6°/sec while the intermittently occluded target moved at one of five speeds centered on 6°/sec. The observer's task was to judge which stimulus moved faster. The duration of the visible and occluded segments were varied systematically and independently between 20–140 msec to create 26 different stimulus conditions. Total duration of all stimuli was set so that the observer saw, on average, 400 msec of motion. Perceived speed and speed discrimination thresholds were computed from the psychometric functions for each visible/occluded duration condition.
The data show that both speed discrimination and perceived speed varied inversely with visible segment duration, but show little relationship with occlusion duration. These results suggest that it is the integrity of the visible local motion signals that determines the accuracy and precision of the speed percept.