Stardust Process: OF

blog_header_experimentsTo create the interactive system for Stardust I needed a tool that would allow me to track people in real time, and then project back the particle system I would create (the stars) on people’s bodies. I had a go at Processing before (for my AR Hockey project) but after discussing it with my tutors, the best option was to use openFrameworks (OF) instead: an open source toolkit that allows you to work with code in order to create interactive work.

OF comes with some examples, but to truly explore its potential you can download addons, or libraries created by the members of the OF community. These addons do specific things, like tracking bodies, for instance. You can adapt and build from these addons into whatever your project needs, and then you can put your project up on a repository (Github, for example) so others can see what you’ve done and build up on it, and so on.

The working ethos of sharing and building up from each other’s work was really helpful for me at stage. I could have never been able to start from scratch, but by using OF I had a play and tried to make sense of the addons I needed in order to materialize my idea.

First, I downloaded OF. To use it, I needed an environment to run it from. As an OS user, I got Xcode, and got the necessary installs (Macports, Homebrew) to get going with OF.

Screen Shot 2013-02-13 at 12.03.09 PMScreen Shot 2013-02-13 at 12.21.07 PMScreen Shot 2013-02-13 at 12.36.16 PM




When everything is installed, it looks something like this:

Screen Shot 2013-02-18 at 10.40.23 AM

I really recommend following the tutorials on the OF site and Github site. There are also a couple of video tutorials that explain step-by step how to get openFrameworks for your computer.

I then started to play around with the examples and addons. I explored the OpenCV addon, and later the ofxCV (the one wrapped by Kyle McDonald), ofxKinect, openNi, among others.

openCVOpenCV blob detection example.

ofxKinect_alphaOfxKinect example.

ofxKinect_pointcloudOfxKinect addon (pointcloud view).

ofxCVTrying an example of OfxCV (unexpected glitch).

ofxCVcontoursOfxCV contours example.

ofxCVflowOfxCV flow example.

openNIOpenNI first try.

openNI2OpenNI. This addon uses the kinect as a depth camera and detects and tracks the body, placing a skeleton and some joints. I used this as a base to create the silhouette as an alpha mask, and place an image of stars “inside” for my screen-tests. 


OpenNI working.

With these tests, then I made my own little app (project) and started making Stardust. I started by making the silhouette (using openNi, ofxCV, ofxKinect), the one that is shown on the full-scale sceen tests. Here is how the first experiments with openFrameworks for stardust turned out:


After, it was iteration after iteration, tweaking and adding code to make a particle system (flocking), make it look like my aesthetic benchmark (like stardust), and adjusting the tracking and of course, the projection!

firma 2



Share your thoughts

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: