Research On 3D Reconstruction Of Shell Parts Based On Stereo Vision

With the rapid development of China's aerospace industry,shell parts have become more and more widely used in this field.Shell parts cannot be directly measured due to the presence of irregular surfaces and hole features.Usually,three-dimensional reconstruction techniques are used to reverse the weight of shell parts.Structure.With the wide application of shell parts,people have higher and higher requirements for the efficiency and accuracy of retrograde reconstruction.This dissertation first analyzes the research status of the three-dimensional reconstruction technology and correlative technology at home and abroad.,laser tracking,and three-dimensional reconstruction of structured light.Traditional contact measurement has limited use due to slow three-dimensional measurement reconstruction techniques,limited range,and inconvenient operation,vision and image processing technologies in recent years.The three-dimensional reconstruction technology based on stereo vision is due to its non-contact type.The measurement method,which does not require special light sources,has strong robustness,low cost,and high efficiency,has been called a research hotspot in recent years and has been widely used.Based on the three-dimensional reconstruction of the shell parts based on stereo vision,this paper develops a threedimensional reconstruction system of the shell parts,and studies and analyzes the key technologies of the system.This dissertation first the research status of 3D reconstruction technology and stereo vision technology at home and abroad are analyzed.The research on 3D reconstruction technology based on binocular vision mainly includes camera calibration and calibration,stereo matching,3D coordinate calculation,marker location and point cloud registration.The calibration part of the camera introduces the principle and method of camera calibration.It uses the Zhang Zhengyou camera calibration method to verify the calibration results of the two platforms Open Cv and MATLAB,and compares the calibration accuracy of the chessboard calibration board and the circular calibration board,finally using Open Cv The anti-symmetric circular calibration plate calibrates the camera's internal and external parameters.The correction section introduces and compares Hartley and Bouguet stereo correction algorithms.The stereo matching part performs stereo matching research on the surface feature points of the housing parts and the edges and feature contours.The sparse point matching algorithm based on the window is used for the edges and feature contours of the shell parts,and the SIFT stereo matching algorithm is used for the surface feature points to complete the stereo matching between the left and right images.,and use a random sampling agreement algorithm to remove false matches.The point cloud registration section uses ICP algorithm to splice local point clouds based on marker points.A three-dimensional reconstruction system of shell parts was developed based on VC++,and a binocular vision hardware platform was set up.The feasibility and accuracy of the system were verified by three-dimensional measurement and reconstruction experiments,and the accuracy was analyzed and verified.The experimental results show that the accuracy of the 3D reconstruction system meets the requirements of 1mm accuracy,and the three-dimensional reconstruction and stitching of the part contour are better.