A 3D rock mass point cloud registration method based on plane polygon matching

Abstract

The invention relates to a 3D rock mass point cloud registration method based on plane polygon matching. The method mainly comprises three stages (1), plane detection based on voxel multi-time regiongrowth; (2), PCA-based polygon matching; and (3), point cloud registration based on matching point pairs. The method comprises the following steps: firstly, dividing initial rock mass point cloud intovoxel grids, and extracting reliable growth units in each voxel by utilizing an area growth algorithm; Performing secondary growth based on the growth unit to obtain an initial plane set; for the initial plane set, remaining point clouds in neighborhoods around each plane grow, and a plane set in the rock mass point clouds is extracted; Secondly, extracting a polygon boundary of the plane point cloud, projecting a polygon to a two-dimensional plane by utilizing PCA, and carrying out polygon matching; the rock mass point cloud registration method based on the ICP algorithm is suitable for theregistration process of large-scene rock laser scanning point cloud, and the rock mass point cloud registration can be efficiently and accurately achieved.

Description

technical field [0001] The invention belongs to the intersecting field of computer vision and three-dimensional point cloud data processing, and relates to technologies such as point cloud processing and three-dimensional reconstruction of rock mass, and particularly relates to a registration method of 3D rock mass point cloud. Background technique [0002] Laser scanners can obtain 3D coordinate information of target objects with high resolution and high accuracy, and are widely used in 3D reconstruction, 3D target recognition, virtual reality and other fields. However, when acquiring a large scene target, it is often difficult to obtain the whole picture of the target in one scan, and multi-station scanning is required. The goal of point cloud registration is to unify the multi-station data acquired by the laser scanner into the same coordinate system. As the initial link of point cloud processing, point cloud registration is a hot research topic at present. [0003] At p...

AI Technical Summary

Problems solved by technology

Moreover, the surface of the rock mass is irregular, rough and of different scales. It is difficult for traditional feature extraction methods to extract effective features from the rock mass point cloud for the registration process.

Images