Binocular Stereo Vision Technology Based On Human Eye Physiological Perceptio

With the development of computer vision,binocular stereo vision technology has gradually become a research hotspot of scholars at home and abroad.More and more stereo vision systems based on the characteristics of the human eye have appeared to observe the world.For example,in virtual and augmented reality and naked-eye 3D applications,specific data sets are displayed.For real-time scene depth acquisition,it will cause dizziness and visual fatigue due to the incompatibility of parallax and device resolution with the human eye.At present,the drone equipment that is truly dizzy-free on the market is basically unavailable.The key technology to solve is stereo matching,and considering human vision,it is still challenging.Starting from theory,this paper uses the proposed algorithm to solve the occlusion and other interference problems of the binocular image.The main contents of this paper are as follows:(1)An improved semi-global stereo matching algorithm based on the physiological characteristics of the human eye is proposed.Starting from the physiological structure of the human eye and the theory of binocular stereo vision,the camera calibration is completed,and the calibration parameters are used for image correction.Analyzing images for possible occlusion,mismatch,noise sensitivity and other issues.Improving the census algorithm to calculate the difference between the center point pixel and the reference point pixel,describing the similarity between the images,and using the multipath energy function is calculated to obtain the minimum cost.Considering that the parallax causes the left and right and front and back occlusions of the image,it is necessary to check the left and right consistency,and eliminate the invalid area,and use the effective parallax to fill it.Finally,filtering is performed to remove noise to improve the parallax effect,and it is verified by the Middlebury dataset,and the algorithm effect is evaluated by the algorithm execution time and the mismatch rate.(2)Build a stereo vision system and verify the reliability of the algorithm through realtime shooting.The selection of software and hardware platforms is in line with the physiological structure of the human eye,and through multiple sets of comparative experiments,the evaluation of the binocular stereo vision system and algorithm has been completed.This paper lays a foundation for the research of small wearable binocular vision products based on the binocular stereo vision based on the physiological perception of the human eye.