论文部分内容阅读
由于缺乏有效并具有针对性的方法以及传统遥感影像的2维局限性,城市立交桥点的识别与轮廓提取存在较大困难。本文基于机载LiDAR点云数据提出一种城市立交桥的提取方法。首先,通过基于局部平面约束的剖面点分割方法,利用多方向扫描线建立改进的剖面邻域点云组织结构,从而提取出立交桥种子点。然后,利用Alphashape算法定位立交桥轮廓,并获取立交桥初始点。最后,利用区域增长的方法提取出完整结构,同时获取立交桥轮廓。实验结果表明,从提取结果上获取的立交桥长度、宽度以及形态等指标与人工量测结果相近,具有较高的准确度。
Due to the lack of effective and targeted methods and the two-dimensional limitations of traditional remote sensing images, the identification and contour extraction of urban overpass points have great difficulties. Based on the airborne LiDAR point cloud data, this paper proposes a method of urban overpass extraction. First of all, based on the method of sectional plane point segmentation based on local plane constraints, the multi-directional scanning lines are used to establish an improved cross-sectional point cloud structure and extract the seed point of the overpass. Then, the Alphashape algorithm is used to locate the overpass profile and obtain the overpass initial point. Finally, using the method of regional growth to extract the complete structure, at the same time obtain the overpass contour. The experimental results show that the length, width and shape of the overpass obtained from the extraction result are similar to the artificial measurement results and have high accuracy.