Complex contour defect verification method based on point cloud three-dimensional reconstruction

Abstract

According to the complex contour defect verification method based on point cloud three-dimensional reconstruction, point cloud preprocessing, registration, curved surface reconstruction and defect verification methods are improved, rapid and fine three-dimensional reconstruction of complex contour point cloud is optimized, and important progress is made in the aspect of defect verification of a complex three-dimensional model. And automatic checking and visualization of the defects of the three-dimensional model are realized. In the aspect of point cloud pre-processing, the improved RANSAC point cloud pre-processing method not only can realize filtering sampling, but also can separate outliers, and has a better processing effect; in the aspect of point cloud registration, through a two-stage registration process based on feature convergence, the registration precision and efficiency are greatly improved; in the aspect of point cloud curved surface reconstruction, the fitting precision is improved, the iteration convergence speed is increased, all indexes are greatly improved, and the effect is better especially for a complex curved surface; in the aspect of defect verification, false defect points are filtered out, and the defect verification precision is further improved.

Description

technical field [0001] The present application relates to a point cloud complex contour defect verification method, in particular to a complex contour defect verification method based on point cloud three-dimensional reconstruction, which belongs to the technical field of point cloud complex contour defect verification. Background technique [0002] As high-precision laser scanning equipment enters the commercial field, point cloud-based defect verification technology has gradually become the mainstream defect detection and quality assurance method in the field of product defect verification. Laser scanning point cloud is used to verify the defects of parts, which has high reliability. It does not need to remove the grease, dust and protective layer on the surface of the parts, and can realize the in-situ defect correction of parts without disassembling and destroying the current structure. test. Therefore, it is of practical significance and great application value to carr...

AI Technical Summary

Problems solved by technology

First, in terms of point cloud pre-processing, the improved RANSAC point cloud pre-processing method can not only achieve filter sampling, but also separate outliers, which has a better processing effect than before the improvement; second, in point cloud In terms of cloud registration, after the two-level registration process based on feature convergence proposed in this application, both registration accuracy and registration efficiency have been greatly improved; third, in terms of point cloud surface reconstruction, this paper The collaborative wide sum method proposed in the application not only improves the fitting accuracy, but also speeds up the convergence speed of iterations. Compared with the existing technology methods, it has greatly improved and improved in various performance indicators, especially for complex surfaces. Better surface reconstruction effect than other methods; Fourth, in terms of defect verification, the distance threshold can better adapt to the change of surface topology, better reflect the characteristics of the area where the points to be verified are located, and improve the identification of defect points Accuracy; at the same time, limit the defect points, eliminate the isolated defect points caused by calculation and three-dimensional reconstruction, make the defect points conform to the regional aggregation characteristics, filter out false defect points, and further improve the accuracy of defect verification

Images