Research On Visualization Method Of 3D Digital Core

Due to the rapid development of computer visualization technology in recent years,in the process of studying underground structures,three-dimensional digital core technology that can simulate the real structure of rock can be reconstructed according to the microscopic pore structure information of rock.The 3D digital core usually has a large amount of data.When it uses traditional 3D visualization calculation,it has a large amount of calculation and low operating efficiency,which increases the research difficulty of researchers on 3D digital core,thus limiting the technology of 3D digital core.Development application.The main research work and its innovation in this paper are reflected in the following aspects:1.Pre-process the 3D digital core data.The three-dimensional digital core is filtered by straight-through filtering combined with the bilateral filtering method to reduce invalid data points.2.Study the modeling principle method of 3D digital core and analyze the internal components of 3D digital core.The physical modeling method of X-ray core CT scanning and the multi-point geostatistical(MPS)numerical modeling method are mainly introduced.Then,the optimal classification number is determined by the gray image of the three-dimensional numerical core slice,and the three-dimensional digital core is segmented by the three-dimensional OTSU method.3.Study various methods of volume rendering in 3D visualization technology.The traditional ray casting algorithm,ray tracing algorithm,footprint table method,miscut deformation method and texture mapping method are mainly studied.Analyze the advantages and disadvantages of these five algorithms and compare the effects.Combined with the accuracy requirements of the 3D digital core,it is determined that the ray casting algorithm works best.An improved ray casting algorithm is proposed,which improves the efficiency of the algorithm compared with the traditional ray casting algorithm.Based on the object space correlation of 3D data volume in digital space,the bounding box technique is used to reduce a large number of invalid light rays.Based on the linear variation principle of neighborhood spatial points,the traditional linear interpolation method is improved by cubic linear interpolation to reduce the operation.Amount;combined with the spatial correlation of points in a three-dimensional data volume,the ray hopping method is used to reduce unnecessary sampling and invalid points in resampling synthesis.5.Programming implements an improved algorithm and visualizes the 3D digital core.The results of the improved image are compared with those of the traditional ray casting algorithm.The paper compares the running time,the image's overall gray value histogram and the same sampling point interpolation results.The results show that the improved ray casting method greatly increases the running speed without significantly degrading the image quality.