基于扫描线点云的飞机蒙皮边界特征提取

Aircraft skin boundary extraction based on scan line point cloud

针对飞机蒙皮类薄壁结构件边界特征提取效率不高的问题,提出一种基于扫描线点云的边界特征提取方法。考虑了测量点分布不均和边界微倒角的影响,在每条扫描线上提取边界特征点,使用边界特征点构造边界特征线。特征点的提取过程包括邻域搜索、测量点位置识别、直线拟合与求交。邻域搜索中将点沿扫描线排序,通过点的离散曲率区分面上点和边界点,使用表面和侧面上的点分别拟合直线,将直线的交点作为特征点。不在现有测量点中寻找特征点的方法,降低了保障特征点精度所需的点云密度且符合设计意图。通过应用实例验证了算法的提取精度和有效性。

To improve the extraction efficiency of aircraft skin＇s boundary feature due to its thin walled structure, an extraction method was proposed based on scan line point cloud, and the influence of points＇ unequal distribution and the edge＇s micro chamfering were considered. Feature points extracted from each scan line were used to construct the boundary feature curve. The extraction of feature points included neighbor search, identifying measured points＇ po- sition and curve fitting. The points were sequenced in the neighbor searching procedure, and surface points and boundary points were distinguished by their discrete curvature. The lines were fitted respectively with points on the surface and side, and the intersection point was taken as the feature point. By extracting feature point beyond the existing measured points, the point cloud density required to ensure the accuracy of feature points was reduced as well as met the design intent. The efficiency and applicability of the method were testified in experiment.