A comparison of binocular depth mechanisms in areas 17 and 18 of the cat visual cortex

Abstract

1. The retinal disparity sensitivity of neurones in areas 17 and 18 of the cat visual cortex was examined. The response of each cell to an optimally oriented slit was measured as disparity was varied orthogonally to the receptive field orientation. Eye movements were monitored with a binocular reference cell simultaneously recorded in area 17 (Hubel & Wiesel, 1970). 2. Two types of disparity‐sensitive cells were found, similar to those observed in the monkey by Poggio & Fischer (1977). The first type, tuned excitatory cells, were usually binocular and had a sharp peak in their disparity—response curve. They responded maximally at the disparity that brought their receptive fields into superposition on the tangent screen. This disparity closely coincided with the disparity at which the reference cell's receptive fields were also superimposed. By analogy with the monkey this point was taken to be the fixation point, or 0°. The second type, near and far cells, were most often monocular. They gave their weakest response (which was usually no response at all) at 0°. On one side of 0° the response grew linearly for up to 4° and then remained at the maximum. On the other side of zero, it remained at the minimum for up to several degrees before rising towards the maximum. 3. The receptive field organization of several disparity‐sensitive cells was examined using the activity profile method of Henry, Bishop & Coombs (1969). The size and strength of the discrete excitatory and inhibitory regions of the receptive fields of a cell could quantitatively account for the shape of its disparity—response curve. 4. The laminar distribution of disparity sensitivity as well as of several other receptive field properties in areas 17 and 18 was studied. The organization of the two areas was remarkably similar in many respects. There was a difference, however, in the proportions of the two types of disparity‐sensitive cells in the two areas. Area 17 contained many more tuned excitatory cells than near and far cells, while area 18 had the reverse distribution. In addition, the cells in area 18 were sensitive to a much broader range of disparities. While both areas contain disparity‐sensitive neurones, these differences suggest that they play different roles in depth vision. 5. Recent psychophysical and neurophysiological evidence has led to a new model of stereopsis in which depth is signalled by the pooled activity of large groups of cells (Richards, 1971). The current results are consistent with this model.

Binocular Disparity and the Perception of Depth

“…Fortunately, a method for determining the responses of binocular complex cells has recently been proposed by Ohzawa et al (1990) based on their quantitative physiological studies (see also Ferster, 1981). These investigators found that a binocular complex cell in the cat primary visual cortex can be simulated by summing up the squared responses of a quadrature pair of simple cells, and the simple cell responses, in turn, can be simulated by adding the visual inputs on their left and right receptive fields (see Figure 5).…”

“…These curves alone are not very useful from a computational point of view because a response can be read from a disparity tuning curve only when the stimulus disparity is already known. We need a quantitative procedure for computing an unknown disparity in a pair of retinal images from the responses of complex cells to the images.Fortunately, a method for determining the responses of binocular complex cells has recently been proposed by Ohzawa et al (1990) based on their quantitative physiological studies (see also Ferster, 1981). These investigators found that a binocular complex cell in the cat primary visual cortex can be simulated by summing up the squared responses of a quadrature pair of simple cells, and the simple cell responses, in turn, can be simulated by adding the visual inputs on their left and right receptive fields (see Figure 5).…”

Why Is Spatial Stereoresolution So Low?

“…Several computational studies have examined matching by correlation (Tyler and Julesz, 1978;Jenkin and Jepson, 1988;Cormack et al, 1991;Kanade and Okutomi, 1994;Weinshall and Malik, 1995). Furthermore, the properties of binocular neurons in the striate cortex suggest that the estimation of disparity begins with signals closely related to binocular correlation measured after the image is filtered by monocular receptive fields (Ferster, 1981 5-8) pointed out the mathematical equivalence between interocular correlation and the disparity energy calculation that underlies binocular interaction in V1 neurons (Ohzawa et al, 1990;Cumming and Parker, 1997). The accompanying paper (Nienborg et al, 2004) suggests that the size of V1 receptive fields limits the ability of single neurons to detect spatial variation in disparity.…”

Evidence for the subsystems in stereopsis: fine and coarse stereopsis

“…In fine stereopsis, subjects can easily fuse the images even when the eye vergence does not change and the magnitude of perceived depth increases linearly as disparity increases, while in coarse stereopsis subjects report frequently diplopia of the two images when stimuli are presented briefly and the magnitude of perceived depth does not increase linearly as disparity increases. Recently, the neurophysiological substrata of the two different subsystems were found (Fernster, 1981;Poggio & Fischer, 1977;Poggio & Talbow, 1981).…”