Estimation method of scattered point cloud data geometric senses

Abstract

The invention relates to an estimation method of scattered point cloud data differential geometric senses, which comprises the following three parts of: estimating an initial normal vector for each point in point cloud data according to the position distributing situation of points in an adjacent field, and carrying out unitization on the initial normal vector of each point; carrying out normal fitting on the initial normal vectors, and computing the normal partial derivative of scattered point cloud at each point; and computing the second-order differential quantities such as the maximum principal curvature, the minimum principal curvature, the principal direction, the Gaussian curvature, the mean curvature and the like of the scattered point cloud at each point through the unitized initial normal vectors and the normal partial derivative. Aiming at the scattered point cloud data with strong noise, the invention can accurately estimate various differential geometric senses of each point and has important application value in computer graphics and computer visual adjacent fields.

Description

technical field [0001] The present invention relates to the technical fields of computer graphics and computer vision, and in particular to a method of obtaining discrete point cloud data in three-dimensional space by laser scanning objects or space scenes, and estimating normal vectors and principal curvatures on such three-dimensional point cloud data. (that is, the maximum principal curvature and the minimum principal curvature), Gaussian curvature, average curvature and principal direction (that is, the direction corresponding to the maximum principal curvature and the minimum principal curvature), the first-order and second-order differential geometric quantities. Background technique [0002] With the development of 3D laser scanning equipment and technology, more and more complex objects and scenes in the real world can establish their 3D discrete point cloud (or sampling point set) data model through laser scanning, and this method based on physical measurement The d...

AI Technical Summary

Problems solved by technology

In addition, the irregular distribution of sampling points and the inevitable noise introduced in the data acquisition process also significantly degrade the performance of some differential component estimation methods that perform well under ideal conditions.

Images