A Refined Extraction Method for Street Trees in Mobile Laser System Point Clouds

Street trees on both sides of the road are one of the important infrastructures of a city. Automatic extraction of street trees is of great significance in the fields of smart cities and smart transportation. Mobile laser system (MLS) enables efficient access to object point clouds around the city roads, providing a high-quality data source for the extraction of street trees. Thus, this paper proposes a refined method for extracting street trees in MLS point clouds data in order to achieve the extraction of the point clouds of a single street tree in the complex urban road environment. First, an identification method for the pole-like parts is designed based on the hierarchical semantic features of the pole-like target. Then an estimating method of point clouds dimension features is designed through analyzing a top-down searching and estimating strategy. The method distinguishes the category by the features of the upper point clouds of the pole-like objects and identifies the artificial pole-like objects and the tree point clouds based on the regional growth based on the category constraint. Finally, to solve the problem of segmenting connected tree crowns, this study designs a segmentation method based on curve fitting of the crown contour, taking into full consideration of the geometric features of the tree crown contour. This method realizes a refined segmentation of connected tree crowns and finally obtains a complete single street tree object. The experimental result shows that the method has a great effect on street tree extraction in different urban road environments.