Research of high-efficiency extraction of edges of large complex components in three-dimensional point clouds

With the rapid development of the automobile manufacturing industry, major automobile manufacturers around the world are stepping up digital transformation to achieve high-flexible, efficient and high-quality automobile production to adapt to the increasingly competitive market situation. Robots play an important role in automobile manufacturing. In the process of automobile manufacturing, the performance and quality of the product largely depend on the size accuracy of the parts, so edge information needs to be tested before the product is produced and assembled. At the same time, in the field of path planning, people need to obtain geometric feature information such as the edge of components in advance. However, large components are due to various the accuracy of the error needs to be scanned on the spot, and then processed after obtaining the point cloud. In this paper, a new method for quickly extracting edge contours from large-scale point clouds is studied. In the process of finding edges in two-dimensional images, some optimizations have also been made to make them more suitable for edge extraction of such plane structures. We also demonstrated the application of this method in a bus and compared it with two traditional methods.