A dual CNN model of cyclopean perception and its application potentials in artificial steropsis

Random-dot stereograms coding 3-D information in their internal correlation are used for probing human stereopsis. Dual cellular neural net (CNN) algorithms that can reveal 3-D surfaces coded in stereograms are reported. The concept of the difference stereogram is introduced and used for coding smooth surfaces. Its importance is due to the fact that difference stereograms of real objects can be created in an optical environment using a projector and camera.<>