Combine Laser Scan Data With Openstreetmap To Produce A Three-dimensional Road Map

With the continuous development of computer technology,the method to acquire spatial data has updated rapidly.Three-dimensional digital map attracts so much attention to be developed.Generating a three-dimensional digital map requires a basic map.Because the Open Street Map(OSM)is open-source and free,it has received widespread attention.However,the height information of the road is very sparse in the OSM,and the mean square error is higher than 5 meters,which makes more and more researchers focus on the generation of highprecision three-dimensional maps.Due to the Light Detection and Ranging(Li DAR)point cloud's high-precision characteristics which the mean square error is about 20 cm,it can extend the OSM to generate high-precision 3D maps.The method of OSM combined with Li DAR point cloud to generate three-dimensional digital map can be divided into three parts: outdoor area processing,indoor area processing and calculation of slope abnormality.The main steps of the outdoor area processing are as follows: the three-dimensional road is projected into a straight line by orthogonal projection,the Hough Transform(HT)is used to extract the set of road candidate points,and the plane is fitted to calculate the height;the indoor channel is mainly based on the associated outdoor channel.The height is obtained by summing the height accoding to their projection distances.For the road with abnormal slope,the weighted Hough Transform(WHT)is used for recalculation.The method is based on the plane fitting,which solves the chanllenge that the road area caused by the aerial view of the airborne laser radar is partially blocked.WHT recalculation solves the problem that the latitude and longitude of the open source map is not accurate enough.It is proposed to project a three-dimensional point cloud onto a two-dimensional plane such that the road is projected into a nearly straight line.Compared with the fitting plane,fitting the straight line with HT improves the efficiency of the algorithm,and the projection automatically filters the interference surfaces such as the roof,and the bus facade.HT can fit multiple lines at a time,which can be well adapted to the existence of multiple roads.The three-dimensional road map was established for the city of Aachen using the airborne lidar point cloud provided by the municipality of Cologne,Germany(using the data set for research only and the error of the data set is approximately 20 cm).The experimental results show that compared with the Ordering Points to Identify The Clustering Structure(OPTICS)method,the height difference of the two algorithms is greater than 1 meter which is accounted for 8.7%.The height difference is less than 1 meter,which is in line with expectations.In the range of height difference greater than 1 meter,it takes up 87% that the proposed algorithm predicts height successfully;compared with the OPTICS algorithm,the algorithm has higher accuracy,and is more robust to the case where the point cloud is occluded partly.At the same time,the method can automatically adjust the neighborhood size when extracting the point cloud.Additionally,it is robust to the change of point cloud density and can be easily extended to other data sets.