Point cloud registration method

Abstract

The invention provides a point cloud registration method. The method comprises the following steps of: obtaining four test feature points, satisfying a preset position, in a test point set and a reference feature point, corresponding to each test feature point, in a reference point set, wherein a test values is determined via Gaussian curvature and average curvature of the test feature points, and a reference value is determined via Gaussian curvature and average curvature of the reference feature points; obtaining a coarse registration matrix which enables distances from the four test feature points in the test point set to the four reference feature points in the reference point set to be minimum; obtaining a coarse registration point set according to the coarse registration matrix and the test point set; and obtaining a precise registration matrix which enables the sum of distances from coarse registration points in the coarse registration point set to a first reference point in the reference point set to be minimum, wherein the distance between the reference point and each coarse registration point in the coarse registration point set is minimum. According to the method, the problems of long time consumption and low precision caused by estimating features of all the data points in existing cloud point registration method are solved, and the calculation efficiency and precision are effectively improved.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a point cloud registration method. Background technique [0002] The point cloud registration method matches the coordinates of one data point cloud (test point set) to the coordinates of another data point cloud (reference point set) through the change matrix between two coordinate systems, where the change matrix contains rotation and translation In order to achieve the consistency of the coordinates of the two point clouds. Since the registration accuracy of the point cloud registration method directly affects the reliability of subsequent processes such as error analysis, in the fields of computer vision and computer-aided geometry such as stereo matching, target recognition, pose estimation and image matching, point cloud registration Method is a critical step. [0003] Existing point cloud registration methods usually include two processes of coarse registration and fin...

AI Technical Summary

Problems solved by technology

[0005] The present invention provides a point cloud registration method to solve the time-consuming and low-precision problems caused by estimating the characteristics of all data points in the existing point cloud registration method

Images