Research Of 3D Reconstruction Algorithm Based On Binocular Vision

Stereo matching algorithm is to find the matching points in two or more images with scene,calculate the disparity values as well as the depth of each pixel,and then realize the three dimensional(3D)reconstruction.It is the hot topic in the field of computer vision.Although many stereo matching algorithms have been proposed,there are still many problems,especially in the regions of depth discontinuities,low texture and uneven illumination.This article analyzes and studies the above problems in detail,and devotes to improve the stereo matching precision and reduce the mismatching rate.Firstly,this article introduces the research background and significance of 3D reconstruction algorithm based on binocular vision,summarizes the development process and research,and deeply discusses the basic principle algorithm.1)The imaging principle and calibration method of binocular cameras.2)The constraints and evaluation criteria of stereo matching algorithm.Besides that,the implementation process of traditional local stereo matching algorithms is analyzed.Secondly,for the problems of high time complexity of stereo matching algorithm based on bilateral filtering,the Adaptive Manifolds Filter(AMF)is used to aggregate initial matching cost,and reduce the time complexity while preserving edge.The matching cost is computed based on color and gradient term to establish the matching cost volume,and then the aggregation process is completed by the adaptive manifolds filtering to keep the image's edges.The background filling method is used to fill the mismatching pixels detected by Left-Right consistency strategy.Moreover,weighted medium filtering is utilized to optimize the filled disparity map.Experimental results show that the proposed algorithm can effectively reduce the time complexity while maintaining high matching accuracy.Finally,for the problems of low matching accuracy in both depth discontinuities and low textured regions of traditional local stereo matching algorithms,an adaptive support-weight stereo matching algorithm based on SIFT descriptors is proposed.First,the original matching cost is calculated with gradient amplitude and gradient phase.Then the adaptive support-window for each pixel is obtained with decision criterion based on the color similarity principle.Finally,with the L1 norm of SIFT descriptors of each pixel,the adaptive support-weight is calculated.Experimental results show that the proposed algorithm can effectively improve the accuracy of stereo matching algorithms in both depth discontinuities and low textured regions,and obtains higher matching accuracy.Moreover,the 3D image is reconstructed based on the disparity map obtained by the proposed stereo matching algorithms,which shows that the obtained disparity map can get better reconstruction effect.