Detection of road ancillary facilities considering MLS point cloud neighboring features
To efficiently and accurately obtain the operation status of road ancillary facilities,a detection meth-od for road ancillary facilities based on the neighborhood characteristics of the point cloud was proposed.First-ly,the index and pseudo coordinate of point clouds were constructed by integrating the neighborhood features of point clouds with the vehicle trajectory.Secondly,a two-phase method was designed to extract lower pole-like objects and complete upper points by utilizing principal component analysis and grid-based search meth-ods,facilitating the detection of pole-like roadside facilities.Then,the extraction of overhead facilities were conducted by combining road boundaries with the elevation benchmarks of vehicle trajectory.The least squares method was used for clearance analysis.The results show that the precision and recall of the pole-like roadside facilities are more than 91％and 90％,respectively.Those of the overhead facilities are more than 93％and 92％,respectively.The calculation time is less than 20 s.All of these can meet the engineering requirements.Due to the obstruction of lower pole-like objects,there are errors in the detection of some pole-like roadside fa-cilities.The road clearance errors calculated by the proposed method are all less than 0.1 m,exhibiting high fea-sibility and accuracy.It can effectively meet the requirements for extraction,detection,and management of an-cillary facilities.
万方数据
维普
