TL;DR: The results indicate that cortical visual neurons are binocularly linked to respond to the relative position and contrast of the images over their receptive fields, and also that both these aspects of binocular stimulation may be utilized by the brain as a source of stereoscopic information.
Abstract: The neural signals in visual cortex associated with positional disparity and contrast texture correlation of binocular images are the subject of this study. We have analyzed the effects of stereoscopically presented luminous bars and of dynamic random-dot patterns on the activity of single neurons in cortical visual areas V1, V2, and V3-V3A of the alert, visually trained rhesus macaque. The interpretation of the results and considerations of possible neural mechanisms led us to recognize 2 functional sets of stereoscopic neurons. (1) A set of neurons, tuned excitatory (T0) or tuned inhibitory (TI), which respond sharply to images of zero or near-zero disparity. Objects at or about the horopter drive the T0 neurons and suppress the TI, while objects nearer and farther have the opposite effects on each type, inhibition of the T0 and excitation of the TI. The activity of these neurons may provide, in a reciprocal way, the definition of the plane of fixation, and the basic reference for binocular single vision and depth discrimination. (2) A second set of neurons includes tuned excitatory at larger crossed or uncrossed disparities (TN/TF) and neurons with reciprocal excitatory and inhibitory disparity sensitivity with cross- over at the horopter (NE/FA). Binocularly uncorrelated image contrast drives these neurons to a maintained level of activity, which shifts, in response to correlated images, toward facilitation or suppression as a function of positional disparity. These neurons may operate in the neural processing leading to stereopsis, both coarse and fine, and also provide signals for the system controlling binocular vergence. These results indicate that cortical visual neurons are binocularly linked to respond to the relative position and contrast of the images over their receptive fields, and also that both these aspects of binocular stimulation may be utilized by the brain as a source of stereoscopic information.
TL;DR: In this study, depth discrimination, using disparity cues alone, was studied with a small fixation point and briefly exposed, vertical slit‐shaped targets.
Abstract: 1. Depth discrimination, using disparity cues alone, was studied with a small fixation point and briefly exposed, vertical slit-shaped targets.2. The upper limit for reliable qualitative localization of a slit as nearer or further than the fixation point is 4-7 deg of absolute disparity in a convergent direction and 9-12 deg in a divergent direction. Even larger absolute disparities can be recognized in the peripheral visual field.3. Relative depth discrimination between two slit targets was measured as a function of their spatial position. The horopter (the locus of targets that appear to be fused binocularly) is the region of maximum stereoacuity and this does not necessarily coincide with the Vieth-Muller circle (the locus of zero geometric or absolute disparity). There is a gradual increase in stereo-threshold as the targets are moved out along the horopter, away from the fixation point into the peripheral visual field. The relative disparity threshold also rises, approximately exponentially, as the targets are moved in depth or absolute disparity away from the horopter.4. Relative depth discrimination is, then, operative over a very wide band of visual space around the horopter (about 3 deg of absolute disparity in the centre of the visual field and even more in the periphery).5. The findings are discussed in relation to the neurophysiology of binocular neurones of the cat cortex. The dimensions of visual space under observation by the binocular apparatus of cat and man are rather similar. The sharper decline of stereo-acuity with absolute disparity in the centre of the visual field may be related to the limits of bilateral representation of a central strip of retina in the human brain.
TL;DR: Both the falloff of sensitivity with disparity pedestal and the disparity range of quantitative stereo depth lead to the conclusion that different size-tuned channels process disparity differently.
Abstract: We have used stimuli with difference-of-Gaussian (DOG) luminance profiles to measure depth-increment thresholds within postulated spatial channels as functions of depth from the fixation plane. Stereoacuity was best with high-frequency DOG’s presented at the fixation plane. Performance was relatively constant for spatial frequency above 2.4 cycles/deg, but it deteriorated as spatial frequency was decreased. Regardless of spatial frequency, stereo sensitivity declined rapidly as stimuli were presented away from the horopter. The falloff occurred more rapidly over the 0–20-arc-min range than over the 20–80-arc-min range. Depth was perceived over a broader range of disparity pedestals with low-spatial-frequency stimuli; however, the lowest thresholds were always obtained with the highest-frequency stimuli. Both the falloff of sensitivity with disparity pedestal and the disparity range of quantitative stereo depth lead to the conclusion that different size-tuned channels process disparity differently.
TL;DR: This book discusses the development of Binocular Vision, as well as alternatives to Fusion and Correspondence, and the role of the Horopter in this development.
Abstract: Chapter 1: Introduction to Binocular Vision: Why Do We Have Two Eyes? Chapter 2: Visual Direction. Chapter 3: Fusion and Correspondence. Chapter 4: The Horopter. Chapter 5: Alternatives to Fusion. Chapter 6: Binocular Summation. Chapter 7: Stereopsis. Chapter 8: Neuroanatomy and Neurophysiology. Chapter 9: Development of Binocular Vision. Appendix A: Viewing the Stereograms In This Book.
TL;DR: The central idea is that localizing attention in 3-D space makes precategorical visual processing sufficient to hold gaze, and the system relies on active control of camera movements and binocular fixation segmentation to locate the target.
Abstract: This article examines the problem of a moving robot tracking a moving object with its cameras, without requiring the ability to recognize the target to distinguish it from distracting surroundings A novel aspect of the approach taken is the use of controlled camera movements to simplify the visual processing necessary to keep the cameras locked on the target A gaze-holding system implemented on a robot's binocular head demonstrates this approach Even while the robot is moving, the cameras are able to track an object that rotates and moves in three dimensions The central idea is that localizing attention in 3-D space makes precategorical visual processing sufficient to hold gaze Visual fixation can help separate the target object from distracting surroundings Converged cameras produce a horopter (surface of zero stereo disparity) in the scene Binocular features with no disparity can be located with a simple filter, showing the object's location in the image Similarly, an object that is being tracked is imaged near the center of the field of view, so spatially localized processing helps concentrate visual attention on the target Instead of requiring a way to recognize the target, the system relies on active control of camera movements and binocular fixation segmentation to locate the target