Fine extraction method of road boundary based on vehicle-mounted point clouds
This paper proposed a road boundary extraction method based on vehicle-mounted point clouds to address the issues of low accuracy and poor efficiency in existing road boundary information extraction.The steps of extracting road boundaries by using this method were as follows:Firstly,to improve the processing efficiency of subsequent algorithms,gradient filtering and cloth simulation filtering(CSF)algorithms were used to separate ground and non-ground points and obtain candidate datasets of road boundary points.Secondly,open street maps(OSMs)were used to assist in segmenting road boundary candidate points,and the road boundary point extraction was achieved through the random sample consensus(RANSAC)algorithm.Finally,the cumulative curvature value and distance of the boundary breakpoint area were calculated to achieve intersection judgment.For breakpoints,a quadratic polynomial curve was used to fill in the fitting.The experiment was conducted by using vehicle-mounted point cloud data.The results show that the accuracy of the proposed method in extracting road boundaries can reach over 80%,which is superior to the comparison method.This verifies the feasibility and superiority of the proposed method,and the research in this paper can provide certain technical references for obtaining urban road information.
vehicle-mounted laser point cloudboundary extractionrandom sampling consensus(RANSAC)cloth simulation filtering(CSF)
万方数据
维普
