A forest point cloud registration method based on trunk position iteration

By using a forest point cloud registration method based on genetic algorithms, the problem of tree canopy and trunk misalignment in forest environments is solved, achieving high-precision automated registration that is suitable for forest resource surveys and ecological environment monitoring.

CN122156264APending Publication Date: 2026-06-05UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
UNIV OF ELECTRONICS SCI & TECH OF CHINA
Filing Date
2026-01-04
Publication Date
2026-06-05

AI Technical Summary

Technical Problem

Existing UAV lidar (ULS) and backpack lidar (BLS) point cloud registration methods suffer from poor robustness and sensitivity to initial pose issues due to significant horizontal offset (vertical misalignment) between tree canopy and trunk in forest environments, making it difficult to achieve high-precision data fusion.

Method used

A forest point cloud registration method based on genetic algorithm is adopted. Through a global search strategy, it is transformed into a continuous two-dimensional spatial distribution map overlap optimization problem. By using tree trunk position iteration and spatial distribution pattern matching, high-precision registration of UAV and backpack LiDAR data is automatically achieved.

Benefits of technology

It achieves high-precision registration without human intervention, overcomes vertical misalignment, and features strong noise resistance and insensitivity to initial errors, making it suitable for high-precision forest resource surveys and ecological environment monitoring.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122156264A_ABST
    Figure CN122156264A_ABST
Patent Text Reader

Abstract

The present application relates to the technical field of remote sensing, in particular to a forest point cloud registration method based on trunk position iteration. The present application makes full use of the natural complementarity of ULS and BLS data in obtaining vertical structure information of forest, converts the traditional discrete point set matching problem into the overlap optimization problem of continuous two-dimensional spatial distribution influence map; by introducing genetic algorithm to construct the spatial mapping relationship between non-homologous data and carry out global optimization, the interference of the horizontal offset of crown and trunk commonly existing in broad-leaved forest on the registration accuracy is fundamentally avoided. The present application can automatically obtain the global optimal solution and generate the homogeneous transformation matrix without laying artificial targets, finally realizes the high-precision rigid alignment of two sets of point clouds, has the advantages of high automation, strong noise-robustness and insensitivity to initial error, and can be widely applied in the fields of high-precision forest resource inventory, digital forestry construction and ecological environment monitoring.
Need to check novelty before this filing date? Find Prior Art