LOSTCIRCUITS

On the last page, the terms Vernier Acuity and binocular disparity were mentioned. Since not everyone is familiar with the terminology, I have made two little cartoons of what this is all about: On the right: Illustration of vernier acuity. Visual accuity can be measured as minimum separabile which means the closest distance that two points or dots are apart from each other before they can no longer be discerned as 2 individual points. For a demo, this is rather impractical, therefore, I am using a horizontal line with a gap of 1 pixel width. Suffice it to say that the presence of the gap is not very obvious. Much better resolution is achieved by the Vernier Acuity, that is, how much offset can there be between two lines until they no longer appear to be one straight line. The vertial lines are offset by a single pixel which appears much more pronounced than the 1 pixel gap in the horizontal line.

Depth perception is mostly relative since the primary cues are relative shifts of the image relative to the optical axis of the eye. That is, objects that are closer are viewed at a greater angle. Therefore, relative to a reference point, the images projected onto the retina fall on the outside (temporally) of the image of the reference point. Likewise, objects further away are projected nasally

It should be clear that binocular disparity can only give results of objects with a relative change in z-depth to each other. Since, however, in most humans, binocular disparity overrides the absolute values of eye position and accommodative state of the lens, it follows that, by extension, relative depth perception overrides absolute depth. This opens the door to a variety of mechanisms to generate illusion of depth perception. Keep in mind, though, that for different humans, the order of importance of the different cues can greatly vary. This means that different humans have different depth perception, which is why some people are virtually blind when it comes to stereo pictures. Nonetheless, almost everyone can be fooled into depth perception if the correct parameters are used.

Next Page:    => Turning 2D into 3D =>