基于视觉追踪技术的三维点云自动拼接方法及精度分析

Automatic stitching method and precision analysis of 3D point cloud based on visual tracking technology

三维点云拼接是三维扫描测量内业处理一个重要环节,拼接效率及精度直接影响后续点云的应用。常用的点云拼接方法需要人工通过控制点、标靶或空间特征点对不同测站的点云进行匹配、拼接。基于视觉追踪技术的三维激光扫描仪的出现,改变了传统三维激光扫描作业模式,外业无需架设标靶或控制点,点云拼接自动完成,无需人工干预,大大提高了作业效率。对基于视觉追踪技术的扫描仪进行测试,对其自动拼接的点云,与人工分别基于空间视图和特征点进行拼接的点云进行精度对比分析,得出结论:基于视觉追踪技术的三维点云自动拼接精度,与基于空间视图和特征点的点云拼接精度一致。

3 D point cloud splicing is an important link in 3 D scanning measurement.The efficiency and precision of stitching directly affect the application of subsequent point clouds.The commonly used point cloud splicing methods need to pass through the control points manually.The point clouds of different stations are matched by target or spatial feature points,and the appearance of 3 D laser scanner based on visual tracking technology changes the traditional 3 D laser scanning operation mode,and the field does not need to set up targets or control points.The point cloud splicing is completed automatically,without manual intervention,which greatly improves the working efficiency.A scanner is used to test the point clouds,which are automatically stitched based on visual tracking technology,and the accuracy of point clouds based on spatial views and feature points is compared and analyzed.It is concluded that the automatic stitching accuracy of 3 D point cloud based on visual tracking technology is the same as that of point cloud mosaic based on spatial view and feature point..