基于八叉树的复杂矿体块段模型高效构建

Rapid and robust creation of block model based on octree from complex orebody model

针对复杂矿体结构模型三角形面片数量大、模型几何质量不高,致使基于结构模型构建块段模型非常耗时、鲁棒性不高等问题,提出通过三维到二维的转换方式,将体元与多面体内外关系的三维判断转化为在二维面上进行多边形区域内外点集筛选的方法,并结合栅格扫描法,实现三维复杂矿体的块段模型快速构建算法;通过引入伪八叉树作为规则网格与八叉树模型的桥梁,实现基于八叉树的块段模型高效构建.实验结果表明:该算法可以处理任意复杂矿体模型构建基于八叉树的块段模型,速度快、鲁棒性高,不仅适用于基于矿体结构模型构建块段模型,也适用于同类三维矢量数据向三维栅格数据的转换.

As there are many problems such as time-consuming, poor robustness when building block model from complex orebody model which has large number of triangles, and its quality of geometry is poor, thus to deal with the problems, a rapid and robust algorithm was proposed by 3D to 2D transformation, as convert the determination of the relationship between the cell and polyhedron to the determination of relationship between point and polygon. Combined with the raster scanning method, fast and robust construction block model was achieved for 3D complex orebody. At the same time, by introducing the fake octree as the bridge between structured grid and octree model, construction of the block model was achieved based on octree. Experiments show that the algorithm can deal with any complex ore body model for building block model based on octree with fast speed and excellent robust- ness, and the algorithm is not only applicable to building the block model of orebody, but is also ap plicable to conversion from 3D vector data to 3D raster data.