| The technology of point cloud processing is developing rapidly along with the rapid development of data measurement technology. The technology to the point as the basis for object reconstruction has the huge advantage for increasing the speed of objects drawn and reconstruction, enhancing the large-scale data processing ability, and increasing the computer processing capacity. It is a hot point in reverse engineering. Along with the social demand for oil, gas and other energy increasing, the geological exploration of the fossil energy such as natural gas and petroleum is becoming more and more demanding. As the development efforts of petroleum exploration increasing and the scale of exploration area expanding, the need to point cloud data processing is growing, and the use of a single processor serial processing method has been increasingly unable to meet the processing of massive point cloud data.In this paper, the problem of the surface reconstruction for large-scale point cloud data has been studied with dividing the original point cloud data into blocks and parallel processing, and the thesis realizes the geological surface reconstruction method for large-scale point cloud data. The main contents of this article are in the following:1. According to the characteristics of point cloud data and geological curved surface, the article builds a set of parallel processing system suitable for complex geological curved surface reconstruction of large-scale point cloud data. The system according to the surface reconstruction process is divided into seven modules: the original point cloud data dynamic partitioning and task allocation module, the original point cloud data in blocks preprocessing module, grid construction and fault projection module, constraint Delauany triangle net construction and space triangulation recovery module, curved surface interpolation module, global splicing and smooth module. Under the control node and compute node coordination treatment, user obtain the smooth reconstruct surface of large-scale point cloud data. The system has been tested on actual project data, through the reconstruction of the spatial curved surface compared with original distribution of the location of the original point cloud data, we verify the practicability of the system.2. According to the distribution characteristics of point cloud and geological constraint rule, an algorithm for geological curved surface interpolation has been proposed. The algorithm according to the original point cloud data distribution scattered, irregularly, and in fault zone exists the same(x, y) coordinates distribution points of different attribute z value, the interpolation data is not accurate easily. The method constructs Delaunay triangle net by the original horizon point cloud data and point constraint of fault, and according to the index of curved surface grid point which needs to interpolate to position the triangle, it uses the interpolation triangle vertex data to interpolate the grid points. Finally, it combines with subsequent surface smoothing algorithm, reduces the surface smooth times, and at the same time, makes the reconstruct surface conforms to the original point cloud data distribution. The algorithm verify its practicability and accuracy through the work area data. |
