Automatic extraction method study of road marking lines based on projection of point clouds

Quick access and updates of urban road network information is an important part of construction and management. With the rapid development of urban cities, traditional surveying and mapping with long operation cycle and slow data updating as its characteristics already cannot meet the requirements. To solve the problem, this paper puts forward an automatic extraction method of road lines based on Vehicle-borne laser scanning point data. First, transforms point clouds into image based on intensity projection; Second, uses some simple methods to process the image, which are erosion, dilation, sharpen edges and binarization; third, extracts road lines in image space by using Hough Transform; finally, searches the points around the road line from point clouds, and then extracts road lines through least squares fitting Method automatically. After a series of experiments, it turns out that road line extraction accuracy is up to 0.04m, which obviously shows that this method has broad application prospects in urban road network database building and updating.