Cylinder feature-based tree trunk point cloud efficient extraction method

Abstract

The invention belongs to the technical field of ground laser radar point cloud data processing, and particularly relates to a tree trunk point cloud efficient extraction method based on cylinder features. According to the tree trunk point cloud extraction method, three-dimensional laser point cloud data in forest vegetation obtained by a ground laser radar is utilized, ground point filtering, normalization layering, tree trunk positioning and tree trunk extraction processing are carried out based on cylindrical features of a tree trunk, and the tree trunk point cloud extraction method for extracting point cloud of a tree trunk below a tree crown from a ground laser radar forest point cloud is established. Compared with an existing tree trunk point cloud extraction method, the tree trunk point cloud extraction method has the advantages that the tree trunk can be accurately positioned, the tree trunk point cloud can be finely extracted, meanwhile, the algorithm efficiency is high, the robustness is high, and the method is suitable for extraction of the tree trunk point cloud in large scenes such as forests and can be applied to many fields such as forestry resource investigation, forest scene reconstruction and forest structure parameter extraction.

Description

technical field [0001] The invention belongs to the technical field of ground laser radar point cloud data processing, in particular to a method for efficiently extracting tree trunk point clouds based on cylindrical features. Background technique [0002] Lidar technology is an active remote sensing technology, which can measure the three-dimensional surface shape of an object without contact, and express it with a large number of coordinate points, that is, point cloud data. Ground LiDAR, including fixed LiDAR, vehicle-mounted LiDAR, and backpack LiDAR, can obtain high-precision, high-density under-forest point cloud data, and provide services for applications such as forest resource investigation, forest scene reconstruction, and forest structure parameter extraction. Efficient and convenient technical means. [0003] A tree can be roughly divided into two parts, including the trunk and the crown. Compared with airborne lidar, ground lidar can obtain detailed tree trunk...

AI Technical Summary

Problems solved by technology

[0006] In view of the above problems or deficiencies, in order to solve the existing tree trunk point cloud extraction methods have low efficiency and low precision (such as under-extraction of the upper trunk, over-extraction of the lower trunk ) problem; the present invention provides a method for efficiently extracting tree trunk point clouds based on cylindrical features, which takes the ground lidar scene point cloud as the processing object, firstly locates the trunk accurately and efficiently through the cylinder fitting and judgment method, and then uses the trunk point Fine and efficient extraction of scene tree trunk point cloud by micro-cylindrical search method

Images