Stereoscopic depth : a biological inspired judgement

Abstract

The present paper aims to contribute to the research ondepth perception mechanisms by using a biologically inspiredmodel of stereoscopic vision. Stereogram images show thathuman beings are able to perceive depth just from thedifferences between the images coming from the retinas of botheyes. The research made aspires at finding support for the ideathat the way stereogram images are perceived by human beingsis just a consequence of the way they see on real worldenvironments; that is, to show that the information received bythe brain for generations is perfectly enough to make usinterpret stereogram images as having differences on depth.Usually, to implement a stereoscopic depth mechanism onrobots two parallel cameras are used. The images supplied bythese cameras are very difficult to match by means of abiologically inspired algorithm since these ones require the useof local information, and the extension of information neededto mach the images from two parallel cameras is too large.Human beings have the ability to move their eyes and whenthey look at a real world object both eyes converge to thatobject. The disparity method built expected to make use of thisbehaviour (eye convergence) in order to reduce the disparitycomputation to local and much simpler algorithms.