Intrepid

This project uses the Point Cloud Library (PCL) to analyze 3D data points and produce a mesh.

Formatting

The input data set is a mx3 matrix where each column represents an x, y, and z coordinate respectively. The input file of points is formatted as follows

x_1 y_1 z_1

x_2 y_2 z_2

...

x_m y_m z_m

with a space between each coordinate of a point on a line.

Subdirectories

build – CMake build files

bin – Executables for Intrepid

data – 3D scans from the Project Tango tablet

Executables

ttp - Converts a .txt file to 2 .pcd files (with/without normals)

ptv - Uses the GreedyTriangulation algorithm to convert the pcd with normals to a vtk mesh

mcr - Uses the Marching Cubes algorithm (Hoppe or RBF) to generate a mesh from a pcd with normals.

intrepid - Takes in a .txt file, performs filtering algorithms, normal estimation, and generates a mesh using Greedy, Marching Cubes Hoppe, or Marching Cubes RBF algorithms. It's the "main" program!

Project Tango

Intrepid usesProject Tango tablets to take 3D scans of environments. The MeshBuilder application was modified to save binary files of the scan data to /sdcard/Android/data/com.projecttango.experiments.unitymeshbuilder/files/, which can be accessible through connecting the tablet to a computer via USB.

In order to convert the binary file to a .txt file with only the depth data, use TangoBinaryDecoder, a Mono C# application.

Marching Cubes

This subproject takes in input .pcd file and generates a mesh that uses either a radial basis function (RBF) or Hoppe marching cubes as described in Surface Reconstruction from Unorganized Points (Hoppe et al.)

Copyrights

Copyright (c) 2015. Mohit Deshpande

Copyright (c) 2012, Willow Garage, Inc.