Stereo Photogrammetry ala Jig Lab

Stereo Photogrammetry software has documented beginnings of more than 20 years ago. It is, quite simply, the calculation of a three dimensional object or space through the measurement of differences between two photos.

Cameras sitting a fixed distance apart sharing a view of the same target take a picture simultaneously. A computer processes those images along with calibration images to create 3D models. Here is an example of a capture rig.

Capture Rig ^^ One of our capture rigs.

Much like human vision, stereo photogrammetry is fraught with innaccuracy but that's changing quickly.

I am honored to work with a group of visionary (pun intended) researchers who formed a company called Thingamajig Laboratories or Jig Lab. Based out of New Zealand, they have created a highly accurate capture system that works with any two cameras in any fixed relationship.

Here is an example of our technology. The following two subject images were captured simultaneously with a projected calibration pattern.

Picture A ^^ Left Camera

Picture B ^^ Right Camera

After processing the model that was generated by the system looks like this.

Screen Capture ^^ A screenshot of the resulting model.

Let's look at this model in motion. This video is a peek at the output point cloud model generated by Jig Lab's system.

With some cleanup this model can be used with a 3D printer to print out a bust of the subject. One point to observe is that this model has been generated using only two images. There are many holes and missing sides.

The following potato chip bag was stitched together using Jig Lab's proprietary algorithms with twelve images 6 pairs taken every 60 degrees.

We've got some more work to complete before we go primetime but the work so far has been very exciting and promising.