A Fast and Progressive Reconstruction Method of Point Cloud Based on Star Hypothesis

Abstract

The invention discloses a fast and progressive point cloud reconstruction method based on a star assumption, which can not only robustly and quickly reconstruct a three-dimensional model that closely surrounds the point cloud, is watertight and has no self-intersection, and has a smooth surface, but also for local addition and deletion of point clouds, Input changes such as moving the center point can also achieve fast reconstruction. The method includes: dynamically collecting the point cloud of the reconstructed model of the reconstructed object; constructing an initial spherical model; projecting the point cloud onto the spherical surface of the initial spherical model to obtain a corresponding projection point, and calculating the grid vertex closest to the projection point ; Calculate the target position of each grid vertex. The initial spherical model is deformed according to the target position to obtain a star-shaped model; construct a star-shaped grid constraint that is smooth and closely surrounds the point cloud; recalculate the coordinates of the grid vertices, and the star-shaped model Further deformation to obtain a star model; continue to input new point clouds, update the center of the star model and continue to rebuild the star model to obtain the reconstruction model of the reconstructed object.

Description

Technical field [0001] Technical Field The present invention relates to geometry processing, and more particularly, but not limited to, one point cloud rapid progressive reconstruction of a star based on the assumption, in order to achieve rapid progressive configuration point cloud data. Background technique [0002] Progressive point cloud reconstruction method is a method for generating a three-dimensional model by the rapid dynamic point cloud data, is simulated for the virtual environment, a medical 3D image reconstruction in the field of mobile robot navigation, and based on the point cloud 3D reconstruction Application. Currently used progressive reconstruction of the main point cloud is divided into three, these three methods and the limitations of the specific content is as follows: First, the implicit reconstruction method, and the need to solve the implicit function to extract the surface mesh, but less so the operation efficiency progressive reconstruction is not suit...

AI Technical Summary

Problems solved by technology

At present, the commonly used progressive point cloud reconstruction methods are mainly divided into three types. The specific contents and limitations of these three methods are as follows: One is the implicit reconstruction method, which needs to solve the implicit function and extract the surface grid, but the calculation efficiency is low, so It is not suitable for progressive reconstruction, and this method usually requires the input of the normal information of the point cloud, and it is difficult to estimate the normal direction during the progressive reconstruction process; the second is the explicit reconstruction method, such as the reconstruction method based on Alpha Shape, it is difficult to maintain the model The surface is smooth and prone to hole structures, and it will take time to improve the parameters if adaptive adjustment is used; the third is the method based on deformation, and uses point cloud data to drive the deformation of the initial mesh model, so that the mesh model closely surrounds the points cloud, but this method cannot avoid the self-intersection phenomenon of grid points and the corresponding numerical problems during operation

In addition, most of the existing methods can only reconstruct the point cloud collected on the surface of the reconstructed object, but cannot be directly applied to the point cloud collected inside the reconstruction

Images