Making Things See

Making a Mesh from the Kinect Data

Now that we’ve learned how to create geometry with Modelbuilder, it’s time to apply that knowledge to our Kinect data. Instead of creating an STL that represents a simple geometrical shape, we’re going to make one that represents the actual physical objects that are visible in front of our Kinect.

In the last section, we chose the positions of our vertices manually, selecting their coordinates so that they’d form triangles that made up the faces of a cube. This time we’re going to create vertices based on the depth points captured by the Kinect. We’ll loop through each depth point and create triangles that connect each one with its neighbors. Each of these triangles will make up one facet of a mesh that matches ...