Surface-Reconstruction-Toolbox

Download latest released binaries: https://github.com/LuigiGiaccari/Surface-Reconstruction-Toolbox/releases/latest

Introduction

This tool was mainly coded during my master thesis. It offers several algorithms to triangulate 3d scattered points for 3d scanning applications. After finding it again on a old hard disk I thought that github could give it a fresh restart. During the thesis, in collaboration with Luca Di angelo and Paolo Di Stefano 3 papers were published

• Luca Di Angelo, Paolo Di Stefano, Luigi Giaccari: A new mesh-growing algorithm for fast surface reconstruction. Computer-Aided Design 43(6): 639-650 (2011). https://pdfs.semanticscholar.org/4332/35ea0ca6b6db86131ceb206bc160aab65614.pdf
• Luca Di Angelo and Luigi Giaccari, 2011, An Efficient Algorithm for the Nearest Neighborhood Search for Point Cloud, IJCSI International Journal of Computer Science Issues, Vol. 8, Issue 5. https://www.ijcsi.org/papers/IJCSI-8-5-1-1-11.pdf
• Luca Di Angelo, Paolo Di Stefano, Luigi Giaccari: A Fast Mesh-Growing Algorithm for Manifold Surface Reconstruction. Computer-Aided Design and Applications Pages 197-220 | Published online: 09 Aug 2013 http://www.improve2011.it/Full_Paper/33.pdf

Description

The toolbox features at the moment 4 main algorithms:

• SCBMesher Fast adv front without Delaunay triangulation support. Can easly generate .5 M trias/secs on a laptop- Based on: A new mesh-growing algorithm for fast surface reconstruction. Computer-Aided Design 43(6): 639-650 (2011)
• RobustCrust Based on the Powercrust always return a watertight surface therefore only supports closed shell-like surfaces
• BallPivoting Simulates a ball rolling on the surface, absed on the paper from Fausto Bernardini.
• QuickTerrain to triangula surface in the z=f(x,y) form, usually called "terrain".

Tools are designed to be used both interactively and from command prompt:

• scbmesher.exe -in bunny.dat –out bunny.stl
• robustcrust.exe –in bunny.dat –out bunny.stl
• ballpivoting.exe –in inputfile –out outputfile -r 1.0
• quickterrain.exe –in inputfile –out output file

All tools uses the -in and -out switch for I/O. The Ball Pivoting also asks for the ball radius.

Input/Output

Input file formats

• .dat (binary) The file starts with an “int” typeindicating the number of points, next data are a list of doubles indicating the coordinates. Points are stored this way: X1,Y1,Z1,X2,Y2,Z2……. These are the Matlab commands to generate such file:

  fid=fopen("filename.dat",'wb');%open input file
  fwrite(fid, size(p,1), 'int');%first data=the numbers of points;
  fwrite(fid,p','double');%points;
  fclose(fid);%close the file

  And this are the C++ ones:

  pFile =fopen("Input.dat", "wb");//open input file
  nwritten=fwrite(&N, sizeof(int), 1, pFile);//first line (the number of points)
  nwritten=fwrite(&p, sizeof(double), N*3, pFile);//Write the points coordinate
  fclose(pFile);//close the file

• .cgo ( ASCII) The first line contains the number of points, the followings the list of 3D coordinates. Coordinates are separted by space, the decimal separator can be a comma or a point. Example:

  6
  0,060649 0,011160 0,059552
  0,060351 0,012167 0,058159
  0,060749 0,012164 0,059571
  0,060267 0,011162 0,058137
  0,060278 0,013168 0,058137
  0,060653 0,013164 0,059557

Output File

• Stl(binary) Standard binary stl file.

Acknowledgements

Many thanks to:

• We used the robust predicates from Jonathan Richard Shewchuk
• Amenta's directory of computational geometry software. We found great ideas in the Power Crust Paper.
• Fausto Bernardini. Our Ball Pivoting implementation is based on his paper.
• Thanks to the Meshlab team for their software
• tetgen http://wias-berlin.de/software/tetgen/ saved a lot of time

Contacts

For bugs, suggestions, questions refer to: giaccariluigi@msn.com Feel free to open issues or pull requests.

Development

Build

Astyle formatting ./Astyle.exe --options=astyle_options.txt ./.cpp ./.h ./*.c