基于LIDAR数据的建筑物自动化重建

Automatic Building Reconstruction Based on LIDAR Data

对基于LIDAR数据的建筑物重建进行研究,提出了一种自动化的建筑物重建方法。根据建筑物的边缘线通常互相垂直或平行这一特点对提取的轮廓线进行规则化。然后在屋顶三角网中随机选取种子三角形进行区域生长,将屋顶分割成不同的平面,通过平面相交得到建筑物的屋脊线。最后通过搜索离建筑物轮廓点最近的LIDAR点云,将搜索到的LIDAR点云高程值赋给该轮廓点。实验结果表明:利用该方法进行建筑物重建具有较高的精度。

Building reconstruction based on LIDAR data was discussed and a new method of automatic building reconstruction was proposed. The extracted contour lines are regularized according to the point that edge lines of buildings are always perpendicular or parallel to each other. Then seed triangle is selected by chance in the roof mesh which grows in the region and the roof is segmented to several planes, aiming to get the ridge lines by the intersection of plains. At last, point cloud nearest to contour point is searched and the elevation of contour point is endowed with the corresponding point. The test was presented to prove the high accuracy of the method proposed.