Main retina information processing pathways modeling

In many fields including digital image processing and artificial retina design, they always confront a balance issue among real-time, accuracy, computing load, power consumption and other factors. It is difficult to achieve an optimal balance among these conflicting requirements. However, human retina can balance these conflicting requirements very well. It can efficiently and economically accomplish almost all the visual tasks. This paper presents a bio-inspired model of the retina, not only to simulate various types of retina cells, but also to simulate complex structure of retina. The Model covers main information processing pathways of retina so that it is much closer to the real retina. In this paper, we did some researches on various characteristics of the retina via large-scale statistical experiments, and further analyzed the relationship between retina's structure and functions. The Model can be used in bionic chip design, physiological assumptions verification, image processing, and a variety of goals.