Research And Implementation Of 3D Reconstruction Algorithm Of Coding Marker Based On Photogrammetry Principle

In order to inspect large free-form surface workpieces in the industrial field,a noncontact,automatic,and efficient three-dimensional robotic scanner is gradually being used.Due to the limitation of the positioning accuracy of the robot,it is often necessary to use a three-dimensional vision measurement system to perform real-time tracking and dynamic measurement of the scanning probe of the robot scanner.Aiming at the problems of occlusion and frame loss that often occur in traditional binocular vision tracking systems,a multiocular vision measurement system is set up based on digital close-range industrial photogrammetry technology,and the current 3D reconstruction algorithms is studied and applied in the system to improve the reconstruction accuracy of the system.The main research results of this thesis are as follows:(1)Based on the theory of digital close-range photogrammetry,a multiocular visionmeasurement system is set up to track and measure the scanning probe of the 3Drobotic scanner;(2)The three-dimensional reconstruction algorithm of the system is improved toimprove the final reconstruction accuracy of the system;(3)In order to reduce the eccentricity error caused by the marker imaging,twoimplementation schemes of the multiocular vision measurement system areproposed based on the existing three-dimensional reconstruction algorithm.Based on the improved algorithm and the proposed system scheme of this thesis,the length of Creaform scale bar is measured by multiocular vision measurement system.The experimental results show that,the deviation between the measurement result and the reference value is less than 0.300 mm,and the standard deviation is not bigger than0.110 mm,which fulfill the application requirements of the 3D robotic scanner.