Research Of Bionic Vision Based On Retinotopic Mapping

Biological vision system can distinguish the target accurately because of its own somecharacteristic, such as the sensibility for moving target and the rapidity for adjustingresolution. The research of neurophysiology showed that the cognitive process of biologicalvision system is an objective and initiative process to get interaction with the outside world,not only a passive reaction. Moreover, Biological vision system has a unique, stable andreliable system through millions of years’ evolution. So studing bionic vision is veryimportant to solve which the traditional computer vision can’t deal such as targetidentification. In addition, the retinotopic mapping mechanism is the important embodimentfor non linear mechanism of the visual system and the physiological characteristicsreflection that vision system can deal with information efficiently. In conclusion, the maincontent of bionic vision system is to establish the cognitive model of the retinotopicmapping and to deal with the moving target based on the model.The aim of the paper is to establish two-dimensional retinal cognitive model based onunderstanding the physiological function of human visual system, to deal background imageby establishing background model and to distiguish the moving target on thetwo-dimensional retinal cognitive model.The first chapter of this paper mainly introduces the physiological mechanism ofretinotopic mapping, researches of the model based on the retinotopic mapping andresearches of bionic vision based on the retinotopic mapping. The second chapterestablishes the2d cognitive model which based on the retinotopic mapping by analyzing theuneven distribution function of the eccentricity and polar angle. The analysis of the threedimensional structures of the visual cortex based on the2d cognitive model is presented.The establishment of the bionic vision cognitive model provides the theoretical researchbasis for target recognition and target tracking algorithm.The third chapter establishes thebackground model and the segmentation algorithm of the background reduced image onbionic vision cognitive model. Then the minimum detection target size under the LPTtransformation is anlysized.The aim of the fourth chapter is to study the biomimetic visualverification application platform. And the practical application model has established.Meanwhile, the bionic vision system principle is also involved. The last chapter summarizesall the researches during the author’s master study and puts forward the research direction inthe future.