A Parallel Computer Preattentive-Vision System for Learning and Extracting Special Features in 2-D Images

A robust parallel multichannel filtering method for modelling preattentive vision is presented in this paper. Our aim is to achieve an artificial perception such that its pertbrmance rather than its working be as close as possible to that of human vision. In establishing a rigorous theoretical basis tbr constructing original and efficient operators to be incorporated in a robust computer vision system, we drew our inspiration from existing computer methods, which mevttably imply approximate functionning, and also from the experimental results on the mechanisms of human vision, which a priori may be considered as optimum. Our model implemtmts a "homothetic-filter bank" (IIFB) where each filter is selectively sensitive to frequency-and-orientation pair just like a visual single cell. This model verifies both the Shannon theorem and what is known of the properties of human vision. The proposed system is independent of rotations and scale changes. Moreover, an appropriate choice of the frequaney-filter function with its parameters allows us to provide tbr parallel prt~essing using ca.~;ade computations.