反求工程中基于边界扩展的三角网格构造

针对非接触式方式测量的大规模散乱点云数据建模，提出一种三角剖分算法，该算法适用于多张自由曲面片构成的曲面物体，尤其适用于含内孔的曲面对象。算法过程包括两个阶段：第一阶段，采用一种空间栅格装点法来进行初始点云数据精简，精简比率通过栅格小正方体单元尺寸控制；第二阶段，构造种子三角形，通过连接已剖分网格区域的边界边与最优扩展点采形成三角网格，从而向外延展，也可以时一个带有内孔的复杂自由曲面直接进行三角剖分，无需人工分区。实验结果表明该算法可以快速、有效地从三维数据点云建立几何模型。

Triangular grid structure in reverse seeking engineering based on boundary expansion

Aimed at the modeling of large-scale sc attered point cloud data in the measurement with non-contact styled manner, a kind of triangulation algorithm was proposed. This algorithm is suitable for curve-surfaced object composed of multi sheet of free curve surfaced patches and especially suitable for target of curved surface containing inner holes. The algorithmic course contains two steps. The first step is to adopt a spatial grid binding method to conduct simplification on the primary point cloud data; the unit dimension of small grid cube is controlling the ratio of simplification. The second step is to construct seed triangle and through the connection of the boundary sides of grid region being triangulated with the optimum expansion point to form the triangular grid so as to extend outward; and may also to carrying out directly the triangulation on the complicated free surfaces with interior holes and need not manual region-division. The result of experiment shows that this algorithm may establish rapidly and effectively the geometric model from 3D data point cloud.