Reconstruction of 3D Meshes from Point Clouds

2/16/00


Click here to start


Table of Contents

Reconstruction of 3D Meshes from Point Clouds

Outline

Problem Statement

Motivation

Applications

Challenges

Surface Reconstruction from Unorganized Points

Goal of surface reconstruction

? noisy and ?-dense

Algorithm

Computing tangent planes

Finding consistent orientation

Euclidian Minimum Spanning Tree

Riemannian Graph

Compute orientation from graph

Obtaining good propagation order

Computing distance function

Extracting isosurface

Collapse edges

Sample results

Graph-Based Surface Reconstruction Using Structures in Scattered Point Sets

Method

Algorithm

1. Compute EMST

2. Extend leaves

3. Recognize ring and path features

4. Extract different parts

5. Connect similar features

6. Connect associated edges

7. Fill wireframe with triangles

Evaluation

A New Voronoi-Based Surface Reconstruction Algorithm

Method

Voronoi vertices and medial axis

Sampling criterion

Creating the “crust”

Voronoi vertices in 3D

Using poles

Evaluation

Classification of methods

Warping

Volume oriented

Volume oriented

Alpha shapes

Summary

Author: Patrick Min