A Three-dimensional Reconstruction Method And Application Of Workpiece Based On Point Cloud Data

The application of 3D reconstruction is very extensive:the manufacturing process can be explored in reverse according to the existing objects to realize the model reproduction of the objects in reverse engineering;in the protection of cultural relics,the models of cultural relics can be stored digitally,which is good for future generations to explore the history.Promoting role.In addition,it has a close relationship with it in the fields of virtual reality,precision medical analysis,the build and planning of the cities and villages,and has important research significance.In this paper,we use the laser scanner to acquire the 3D information of the work piece.The bilateral filtering algorithm is used to process the point cloud data of the work piece,which can not only availably eliminate the noise,but also keep the detailed characteristics information of the work piece.In the point cloud registration stage,sampling is performed first,the fast point-specific histogram features of the data are extracted,the sampling consistency is roughly registered,and the latest iterative algorithm is performed for fine registration to obtain the complete point cloud data of the work piece.For the overlap or approximate overlap of the spliced point clouds,the redundant points are removed and the point clouds are homogenized to improve the normal vector estimation accuracy of the point clouds,which not only retains the original features,but also reduces the number of point clouds and speeds up Reconstruction speed.The current commonly used surface reconstruction algorithms are analyzed and compared,Poisson reconstruction algorithm,Moving cube reconstruction algorithm and Greedy triangle reconstruction algorithm.The surface after Poisson reconstruction is smoother,and the reconstructed model is watertight,but for non-closed data,the reconstructed surface will generate redundant redundant surfaces.Moving cube reconstruction can't make the local grid evenly distributed,the surface has larger holes,and the detailed features show poorly.This time,the greedy reconstruction algorithm is used to reconstruct the three-dimensional work piece,which can make the surface of the work piece smooth,describe the surface features of the work piece more delicately,and make the reconstruction more accurate.Use Geomagic Control software to match the reconstruction model of the work piece with the standard model using the best fit function.Then compare 3D and 2D to calculate the overall deviation and cross-sectional deviation.According to the analysis of the comparison result,the error meets the work piece design standard.