The Preprocessing On The Point Cloud Of Large And Complex Mold

The mold is an important lacing fixture in the process of mechanical manufacturing. The wear or damage of the large and complex mold, which was called as die failure, will cause huge economic losses, the rapid digital repair and remanufacturing for these molds can bring a huge industry profits. Currently, with the development of optical scanning equipment, especially the technical development of the laser scanner, the point cloud that we collected is becoming more intensive from large and complex mold, the scanned point cloud generally have reached the one million or one hundred million. For the existing ordinary computer processing power, many difficulties will be encounted in the follow-up to the point cloud data modeling, so, before carrying out the model reconstruction, necessary pretreatment should be done to the dense point cloud data.From the study of existing point cloud pre-processing algorithms, based on the large and dense point cloud features obtained from the laser scanner, the point cloud denoising and simplification is respectively studied, and under the Windows operating system, by using Visual C++and OpenGL to build a software platform for the algorithm to achieve the effect, the effect of pretreatment on the point cloud data is good.Search for point in scattered point cloud from large molds is a time-consuming work in point cloud pretreatment, because the k-nearest neighborhood is used to compute the normals, mollification and denoising prerequisite. Using the octree data model to establish the spatial data index, and then a fast k-nearest neighborhood algorithm is used to reduce the setup time, and also improved the computational efficiency. Based on the k-neighborhood relations, we used a neighborhood average algorithm to do the denosing. The point cloud simplification algorithm based on the shape of the discrete operator is adapted to large and dense point cloud, due to its sensitivity to the changes of the model, it can not only maintain the details of the characteristics of the original model, but also can greatly reduce redundant data points, while the computation time required is less. Through the research on point cloud pre-processing algorithm, using the leading performance of OpenGL library in Visual C++environment, software platform was built for the basic graphical interactive operation, through the tests of the algorithm for a large and dense point cloud, the pretreatment process has a good application value.