Ground and airborne LiDAR point cloud registration method based on single-tree segmentation

Abstract

The invention belongs to the field of forest resource operation management and dynamic monitoring, and discloses a ground and airborne LiDAR point cloud registration method based on single-tree segmentation, and the method comprises the following steps: 1, carrying out the LiDAR point cloud preprocessing; 2, obtaining a two-dimensional elevation projection image; 3, detecting single tree data by utilizing a Hough transformation circle detection method; 4, utilizing a seed region growing method to obtain a single-tree-point image; 5, converting the single wood point image obtained in the step 4into corresponding three-dimensional point cloud data; 6, calculating ICP registration parameter based on a single tree center; and 7, carrying out registration, and carrying out precision evaluationon a registration result. In the point cloud registration process, an ICP algorithm based on individual tree segmentation is adopted for fine registration, the number of point clouds used for calculating registration parameters is reduced, and the efficiency of the algorithm is greatly improved on the premise that the precision is guaranteed.

Description

technical field [0001] The invention belongs to the technical field of forest resource management and dynamic monitoring, and in particular relates to a ground and airborne LiDAR point cloud registration method based on single tree segmentation. Background technique [0002] In recent years, LiDAR technology has been applied in forestry monitoring due to its characteristics of direct, fast and efficient acquisition of three-dimensional spatial data, wide monitoring range, high precision, and low operating cost. Due to the occlusion between ground objects and the limitation of the platform position, neither the ground laser scanning technology nor the airborne laser scanning technology can complete the complete scan of the tree independently, and can only scan to the bottom or top of the tree respectively. Both have great limitations, but through the processing of point cloud data, the advantages of the two can be complemented to obtain more complete tree point cloud data, wh...

AI Technical Summary

Problems solved by technology

The ICP algorithm was first proposed by Besl and McKay. This is a registration algorithm based on free-form surfaces. Its advantages are simple and easy to implement, but it is not efficient, robust and sensitive to the initial iteration value.

Images