Research On 3D Reconstruction Method Of Workpiece Based On Structured Light Stereo Vision

With the rapid development of intelligent manufacturing,three-dimensional measurement and reconstruction technology as one of the means of machine perception of the world is more and more attention by the advanced manufacturing industry,especially in the process of industrial workpiece manufacturing,sorting and assembly,three-dimensional reconstruction technology based on structured light stereo vision can not only provide three-dimensional information of the workpiece,but also analyze and process on the computer to get the digital model of its surface,which provides the basis for the subsequent display,modification,analysis and production of industrial solid models.The3 D reconstruction technology based on structured light stereo vision not only provides the3 D information of the workpiece,but also can analyze and process it on the computer to obtain the digital model of its surface,which provides the basis for subsequent display,modification,analysis and production of industrial solid models.In this paper,we take the workpiece in industrial manufacturing as the research object,investigate the process of measuring and reconstructing the 3D data of the workpiece,combine the structured light technology and binocular stereo vision technology,and propose a 3D reconstruction scheme of the workpiece based on structured light stereo vision to realize the 3D measurement and reconstruction of the surface of the workpiece in the industrial site,recover the 3D data of the surface and construct a digital model of the surface under the current viewpoint.The main research contents are as follows:Firstly,the structured light coding and raster image pre-processing techniques are studied.Complementary Gray code technique is selected as the coding scheme to generate raster images,and the scene images with workpiece surface information are obtained by joint trigger control of projection and acquisition,and the captured scene raster images are pre-processed to complete the background removal and differentiation of bright and dark light stripes.Secondly,the structured light decoding and sub-pixel optical stripe edge extraction methods are investigated.Before decoding the pre-processed series of raster images,a sub-pixel positioning algorithm based on complementary curve fitting is used to extract the edges of structured light stripes,and sub-pixel optimization is performed to obtain more accurate stripe edge information,and then the sub-pixel edges are decoded,which reduces the decoding error and improves the positioning accuracy.Then,the 3D data measurement,processing and reconstruction techniques are investigated.Firstly,according to the principle of binocular stereo measurement,epipolar line correction was performed on the left and right cameras,and the corresponding pixel points of the left and right cameras were stereo matched,parallax calculated and the 3D data of the scene were recovered by the matching technique based on structured light stereo vision.Then the 3D data of the scene is pre-processed to obtain the 3D data of the target workpiece,and finally it is reconstructed by the greedy projection triangulation algorithm to obtain the digital model of the workpiece surface.Finally,the design and experiments of the 3D reconstruction system are completed.Firstly,the overall scheme design of the workpiece 3D reconstruction system based on structured light stereo vision is carried out,including hardware platform construction and software system design.Then the optimal working distance of the system and the overall measurement accuracy of 3D data are determined,and finally the surface 3D reconstruction of different workpieces is carried out to analyze the effect and applicability of the system 3D reconstruction.