(The title for this post is a quote from a long-ago Migraine Boy strip.)
On something of a whim, after I finished teaching my final course for the semester this week, I purchased a Microsoft Kinect “game controller.”
I am not someone who plays computer games, and I do not own a Microsoft XBox gaming console. But after becoming peripherally aware of some experiments in motion capture and computer interface control that made use of the Kinect, I decided to see if I could come up with a “Lewis use” for one of these things…a “Lewis use” being one that lends itself to the production of imaginary large-scale potentially-inhabitable abstract sculptures some people consider architecture (I certainly don’t fool myself).
And Kinects were on sale, and I had some store credits…if I must further justify this minor extravagance.
3D scanners were astronomically expensive when I used one while in architecture school, less than a decade ago…the price of this little piece of gaming hardware barely ran into three digits, once I applied my discounts.
It turned out to be a trivial “technical” task, compared to others I have undertaken, to “hook up” this toy to my office computer and find some opensource drivers and other software that would allow me to make something of the information it captures. Someone else has already done all of the heavy lifting, and I am not sure if there is any place here where I need to apply my pathetic computer coding skills.
Of course, the first subject for my 3D-scanning experimentation was myself.
These remind me of the illustrations used for science-fiction paperback covers in the 1970’s. And I do see a great deal of Lewis-use potential here!
For the interested, the best description of the process of attaching a Kinect to a Windows PC that I have discovered is here (see the sidebar link for “Installation instructions”), and the same gentleman, Japser Brekelmans, also provides freely a quite capable piece of software called Brekel Kinect (see the sidebar link for “Download”) that seems to meet my initial needs as I investigate the capabilities of this re-purposed trinket. Mr. Brekelmans’ software is aimed at “motion capture” for the film, television, and gaming development, but it does a fine job converting the Kinect’s data stream into three-dimensional models that will presumably work with my architectural applications.
The self-portrait images above are renderings from a three-dimensional model exported as an OBJ file from Brekel Kinect; cleaned, corrected, and simplified (slightly, using quadric edge decimation collapse) with the open-source application Meshlab;
and then imported into Google SketchUp. Brekel Kinect also outputs a pair of image files with the model:
I used the RGB output to texture two “of myself” in the quadruple-self portrait above (notice the elf peeking over my shoulder!). I can see the depth image type of output being useful for architectural illustration development.
Interestingly enough, SketchUp (unlike Rhino or Blender, the other applications I tested with the Kinect-output) had the most problems with the Brekel Kinect model, as it is imports only obsolete and limited model file formats (3DS) or completely unreliable ones (DAE) …this flaw may in the end be what drives me to move to other modeling applications, for future “Lewis use.” The original model required fairly extreme “tweaking” in MeshLab before I could import it into SketchUp successfully, but opened right away with the other titles.
The finished images above are of course output from SketchUp that were masked, layered, and otherwise manipulated with Adobe Photoshop, per my usual process described in previous posts.