Claim details
Research claim
Disparity detector uses a complex cell architecture sensitive to 56 orientation and disparity combinations, crucial for accurate depth perception.
Computer Vision
Eye Movements (Psychology)
Stereo Vision (Computer Vision)
format_ink_highlighterQuoted evidence
"Disparity detector population module Disparity information can be extracted from a stereo image pair by using a distributed cortical architecture that resorts to a population of simple and complex cells. The population is composed of cells sensitive to N p ×N o vector disparities with N p = 7 magnitude values distributed in the range [-Δ;+Δ] pixels and along N o = 8 orientations uniformly distributed between 0 and π (see Fig. ). The response of each complex cell r c is modeled by the sum of the squared response of a quadrature pair of simple cells (see Fig. ) . The output consists of 56 (7×8) maps of the same resolution as input stereo image."