Week 3: 3D scanning and printing

Goal this week is to digitally combine the glasses CAD design that I made in week #1 with a 3D scan of my head, and then print out a 3D object of myself wearing the digital glasses. Note that I cannot scan these glasses since they don't yet physically exist.

Downloaded ModelBuilder for Processing: library

To interface with the Kinect and generate the mesh from the Kinect point cloud, I used this Processing sketch from the excellent book [Making Things See by Greg Borenstein (O’Reilly). Copyright 2012 Greg Borenstein, 978-1-449-30707-3], with tiny modification for the new version of the library: code

Point cloud from Kinect:


Importing STL into Meshlab. Depth too high:


Reasonable depth:


Reducing mesh density:


Adding a layer from the glasses CAD file:


Relative translation of layers:


After translation and scaling:

Layers were then flattened to create a single STL mesh.

Importing, centering and scaling in replicatorG:


Gave errors when building gCode. So to be safe I went back to the no-glasses scan.

This was able to generate Gcode in replicatorG without errors. So I sent it for 3D printing. Looks like it worked!