Depth from shading and disparity in humans and monkeys

Abstract

This study compared the effectiveness of binocular disparity and shading cues for the perception of depth in humans and monkeys. Using a novel display, we presented shading and disparity cues conjointly, separately, and in conflict with each other in an oddities depth discrimination task. An example of the figures used in the display can be viewed in the form of an autostereogram at http://web.mit.edu/bcs/schillerlab/vss05abstract.htm. The following is a summary of our findings:

1. Both shading and disparity cues are effectively utilized for the perception of depth by humans as well as by monkeys.

2. Depth information can be processed very rapidly when shading and disparity cues are provided conjointly. When presented singly, the processing of disparity cues takes significantly longer than the processing of shading cues.

3. By placing shading and disparity cues in conflict with each other we assessed their relative effectiveness. Processing times were significantly increased under conflict conditions suggesting that normally these cues are co-processed. Similar values were obtained in humans and monkeys.

4. Control conditions for shading included the rotation of the display by 90 degrees thereby reducing the effectiveness of differential luminance values as a depth cue. Performance under these conditions became significantly worse and reaction times were increased.