Retinal neural network model for selective fixation in machine vision

Fixation is the process of looking at a target object using eye movements and occurs not at random but selectively to the particular visual stimuli. Because it plays a great role not only in human vision but also in computer vision, it is necessary to study selective fixation more for industrial implementation. In this study, we propose a selective fixation model which is partially based on the anatomical features in the retinal neural network, and show how it emulates the human fixation.