Automatic extraction method of unstructured road boundary line based on vehicle laser point cloud

Abstract

The invention provides a method for automatically extracting unstructured road boundary lines based on vehicle-mounted laser point clouds, including scanning line segmentation, topological relationship network construction, using the moving window method to obtain ground points, and performing clustering through the K-nearest neighbor clustering method. Eliminate pseudo-ground point clusters based on the number of laser points in the point clusters to obtain road surface laser point cloud data; the outer contour point of road surface laser point cloud data is set as the position of the road boundary line, and the outer contour boundary of each scan line is Point extraction, take the two farthest ground points in each scan line, and finally obtain the outer contour points on both sides of the road; perform cluster analysis to obtain the left and right contour point sets of the road; boundary line tracking Optimize to obtain the optimal boundary line. This patent directly realizes the extraction of unstructured roads and road boundaries directly based on laser point cloud data, provides an effective solution for the automatic extraction of unstructured road boundaries, and improves the degree of automatic extraction of road boundaries.

Description

technical field [0001] The invention belongs to the field of surveying and mapping geographic information construction, and in particular relates to a method for automatically extracting unstructured road boundary lines based on vehicle-mounted laser point clouds. Background technique [0002] Road boundary information is one of the important components of the current surveying and mapping geographic information construction. Its fine, accurate, efficient and fast automatic extraction technology is very important for new technologies, new directions and new fields such as unmanned driving, smart transportation, smart navigation and 3D maps. The rapid development of The automatic extraction of road boundary lines includes two types, the first is the automatic extraction of structured road boundary lines, and the second is the automatic extraction of unstructured road boundary lines. For the automatic extraction of the first road boundary line, directly use the curbs or curbs...

AI Technical Summary

Problems solved by technology

The method of extracting feature images based on point cloud conversion has a loss of accuracy, because in the process of converting a 3D point cloud into a 2D image, there is inevitably data interpolation, and the interpolation process directly leads to the loss and reduction of data accuracy

At present, there is no automatic extraction of unstructured road boundary lines directly using laser point cloud data.

Images