Visual volume calculation in 3D space basically is based on mesh model nowadays which is constructed from scattered point cloud. It exposes inferiority like huge calculation and low speed when volume is needed only. According to that,an algorithm of rapid volume calculation was proposed. First,the convex hull of point cloud was subdivided into tetrahedron with improved incremental Delaunay triangulation. Second, fitting quadric surface and MST of points with KNN were calculated to get normal vectors. Third,those tetrahedron in vitro were removed by using normal vectors. Finally, all tetrahedron's volume were added up to get object's volume. Experimental results show that this method can improve efficiency greatly compared with common methods and keep precision of volume calculation.%三维可视化体积计算基本上都是先由散乱点云构建出表面网格模型,然后基于网格模型计算体积,存在计算量大、速度慢的缺点.针对此问题提出一种快速体积计算法,首先使用改进的增量式Delaunay三角剖分对散乱点云进行四面体剖分;然后利用K近邻计算散乱点的拟合曲面和最小生成树,得到各点的法向量;由各点法向量别除体外四面体;最后计算各四面体体积之和从而得到总体积.实验表明,该算法不仅保证了计算准确度,而且较传统算法大大提高了效率.
展开▼
