Study On Point Cloud Orientation For Three Dimensional Laser Scanning Based On Bundle Block Net Adjustment Method

The rapid development of three laser scanning technology makes its way of surveying a high speed, high accuracy and wide range model. The high density and resolution of scanned point cloud data put forward higher requirements for the processing of 3D laser scanning data.Due to the limited field of view of laser scanners and the complexity of local environment, a single scan can’t acquire full data of the object. And therefore, point clouds are typically collected from different view points via multiple scans. Then the alignment of these data sets, generated from different scan stations, should be united into a common coordinate system. This procedure is known as point cloud registration or point cloud orientation.This paper proposed a point cloud orientation method based on buddle block net adjustment to improve the poor stability, insufficient accuracy by long distance orientation, which can be achieved by follow steps.1. Build buddle block net with beams between scan stations and targets at whole site.2. Set common adjustment equations according to three dimensional coordinate transformation model with the condition that beams toward at the same target are intersected.3. Calculate orientation parameters of each point cloud and determine orientation matrixes as a whole to complete the procedure of point clouds orientation.Finally, comparisons between existing methods and our method are made using the precision of absolute position and relative position for scanned points and the results show a consistency that point cloud orientation method based on buddle block adjustment are of higher precision and stability.