Accession Number : ADA259892
Title : Why Do We See Three-Dimensional Objects
Descriptive Note : Memorandum rept.
Corporate Author : MASSACHUSETTS INST OF TECH CAMBRIDGE ARTIFICIAL INTELLIGENCE LAB
Personal Author(s) : Marill, Thomas
Report Date : Jun 1992
Pagination or Media Count : 33
Abstract : When we look at certain line-drawings, we see three-dimensional objects. The question is why. Why not just see two-dimensional images? Our theory is that we see the objects rather than the images because the objects are simpler than the images. We define the complexity of an object as the number of bits in a pose-independent, binary representation of that object. We examine a number of examples and find that in each case the seen object is indeed simpler than the given image. This leads us to our second question. Given that we are going to see a three-dimensional object when we look at a line-drawing, which three-dimensional object will we see? Our theory is that the vision system will pick the simplest object from among the infinite set of possibilities. We examine a number of examples and find that in each case the data is consistent with the theory. This work is based on the pioneering ideas of Solomonoff and Kolmogorov, and on the more recent minimum description length concepts of Rissanen.... Vision, Three-dimensional, Perception.
Descriptors : *IMAGES , *VISUAL PERCEPTION , *DRAWING(FORMING) , TWO DIMENSIONAL , THEORY , VISION , THREE DIMENSIONAL , LENGTH
Subject Categories : Psychology
Anatomy and Physiology
Distribution Statement : APPROVED FOR PUBLIC RELEASE