A Bio-Inspired Contrast Adaptation Model and Its Application for Automatic Lane Marks Detection

Abstract

Even in significant light intensity fluctuations human beings still can sharply perceive the surrounding world under various light conditions: from starlight to sunlight. This process starts in the retina, a tiny tissue of a quarter of a millimeter thick. Based on retinal processing principles, a bio-inspired computational model for online contrast adaptation is presented. The proposed method is developed with the help of the fuzzy theory and corresponds to the models of the retinal layers, their interconnections and intercommunications, which have been described by neurobiologists. The retinal model has been coupled in the successive stage with the Hough transformation in order to create a robust lane marks detection system. The performance of the system has been evaluated with the number of test sets and showed good results.