Binocular Depth Perception of “Julesz Patterns” viewed as Perfectly Stabilized Retinal Images

Abstract

THE role of eye movements in the promotion, and maintenance of binocular fusion and depth perception is not clear, although experiments^1–3 suggest that judgements of depth in a stereoscopic situation may be made with some accuracy in the absence of eye movements. Langlands found reliable perception of depth in normal vision with short exposures of the order of 10⁻⁵ sec, but it seems uncertain whether judgements could have been influenced by after-images. It should be possible in principle to solve this problem by “stabilizing” images on the retina after the method of Ditchburn⁴ and Riggs⁵, and in 1963 one of us (C. R. E.) tried to study binocular vision with a contact lens stabilizing system in each eye, but with inconclusive results. Partial destabilization because of poor contact lens fit was a probable cause of this lack of success, as Barlow pointed out in a general criticism of this method⁶. Experiments with an after-image as a “perfectly stabilized image” have been undertaken here⁷, and, because with suitable methods prolonged clear after-images of patterns can be obtained, we decided to try to study complex stereoscopic patterns in these conditions. (Because the after-image is formed as the result of temporary changes in the state of the retinal cells themselves, it can be considered to be a completely “stabilized” image.)