Biological systems, including humans and other animals, seem to be able
derive, with apparent ease, a great deal of useful information about their
surroundings from the interpretation of visual images. Computer-based
systems, on the other hand, have not, so far, been particularly successful
in solving this type of problem.
The seminar will briefly review some of the features of the visual systems
of mammals, with particular reference to primates (including humans), and
describe attempts to simulate these in a biologically plausible
manner. Some success in replicating a number of functional properties of
"early" vision has been achieved, including the detection of simple
feature elements and the derivation of information about the relative
disparity of features from the analysis of stereo pairs of images.