Abstract
Using a random-dot display that makes it possible to provide disparity and motion parallax cues separately or in combination, we examined how effectively these two cues can be processed by monkeys and humans when contrast, size, and spatial separation between target and background are systematically varied. Monkeys and humans were trained to first fixate a central fixation spot and to then make a saccadic eye movement to the target that appeared either singly in one of four locations (detection task) or simultaneously with three identical distractors (discrimination task). In monkeys correct choice was rewarded with a drop of apple juice. Our results show the following: (1) Providing both disparity and parallax cues yielded higher percent correct performance and shorter saccadic latencies than when disparity or motion parallax cues were provided singly. (2) Reducing the size and contrast of the display reduced performance more dramatically for stereopsis than for motion parallax. (3) As spatial separation between the background and the target was increased, performance fell off more dramatically for motion parallax than for stereopsis. (4) When the stimuli were presented at isoluminance, the deficit was more pronounced when the display was blue/yellow than when it was red/green.
Our findings suggest that depth based on disparity cues is processed preferentially by the midget system, whereas motion parallax cues are processed preferentially by the parasol system. The koniocellular system that processes color information in the blue/yellow domain appears to make a limited contribution to depth perception.