Research On Surface Reconstruction Based On Vision Measurement

In view of the disadvantages of the traditional forward engineering manufacturing technology,which is tedious and takes a long time,reverse engineering manufacturing technology has become a research hotspot.The key technologies include data collection technology,data processing technology and surface reconstruction technology.Therefore,based on reading a large number of literature,the following contents are studied:1)In this paper,the line structured light platform is used for data acquisition.Firstly,the principle of line structure data acquisition is analyzed and the model of the equipment used is determined.Then,the parameters of the line structure platform are calibrated.Finally,the characteristics of the line structure light strip are analyzed.After comparing four kinds of light strip denoising algorithms with different weight Gaussian filtering algorithms,the light strip denoising algorithm with weight 0.5 is selected.2)After getting the point cloud data,the point cloud noise data processing algorithm is compared and analyzed,and Gauss filter is selected by the subject to denoise the point cloud data;in the comparison of the simplified algorithm,according to the data characteristics,the non-uniform network algorithm is applied to simplify the point cloud data,which makes full preparation for the subsequent reconstruction work.3)After the point cloud data processing,the surface reconstruction is also the focus of the task.The helmet is selected as the experimental object.After the uniform simplification processing,polygon network processing and point line surface processing,the Berzier surface algorithm is applied to complete the reconstruction experiment,and the requirements of the fitting degree and light order are met.4)The hard disk is chosen as the experimental object,and the octree algorithm is used to index the data before the reconstruction,and the reconstruction of hard disk is completed with high efficiency and high precision combined with the NURBS surface reconstruction algorithm.The reconstruction error is less than 0.05 mm,and the reconstruction time is greatly reduced compared with the traditional reconstruction algorithm.Figure41;Table9;Reference 53.