Week 5: Stereovision

It demonstrates how 'magic eye' images or stereograms exploit binocular disparity to create artificial depth, encouraging viewers to focus beyond the screen to reveal a 3D image.

The video explains how early 3D films used red and cyan lenses to filter images to each eye, reproducing binocular disparity. It details the careful choice of colors to maintain full color perception. It then discusses modern polarized 3D glasses and active shutter glasses, highlighting how they achieve similar depth perception by presenting different images to each eye.

The concept of corresponding points on the retina is introduced, where objects directly in focus fall. The video then defines the horopter as an imaginary arc containing objects that are the same distance from the viewer as the fixated object, and thus also fall on corresponding retinal points.

It explains that objects outside the horopter fall on non-corresponding retinal points, leading to a phenomenon where these objects may appear doubled. An example is given by focusing on a distant object while holding a finger in front.

The video clarifies that the horopter is not a fixed area but shifts based on the point of fixation. This shift influences the type of disparity generated by other objects: crossed disparity (for objects closer than fixation) and uncrossed disparity (for objects farther than fixation).

The processing of correspondence and disparity cues in the visual cortex is discussed. Binocular cells in the cortex are responsible for discerning object depth, with some tuned to corresponding retinal points and others to non-corresponding points (visual disparities).

The video introduces binocular disparity as a crucial depth cue, explaining how our two eyes perceive slightly different images, which the brain then stitches together to create a cohesive perception of depth.

The video concludes by explaining that depth cues are rarely used in isolation and can be combined to create powerful illusions. The Ames room is used as an example, showcasing how manipulated visual cues like perceived rectangularity, floor level, and texture gradient combine to create the illusion of differing person sizes.