Building 3D augmented reality markers
- Motivating the problem: Daqri Elements 4D
Augmented reality (AR) markers are typically ugly black and white 2D printouts that serve as placeholders for 3D objects when viewed through an AR display. Immediately, we see a mismatch: a 2D marker represents what is actually supposed to be a 3D object. Suppose you want to move an virtual horse that is projected onto such a marker. Reaching for it, your hand would move through thin air. There is no 3D object to interact with. Instead, you have to find the 2D paper and move it.
Real world objects have incredible tactile richness. As you pick it up, you immediately feel its weight, texture, rigidity, temperature, etc. Depending on the object, you can also manipulate it in interesting ways too. Even a simple 3D object like a die can be rolled or placed on any side. There are some examples of 3D AR markers. Daqri Elements 4D is a nice one. In their application, each side of a cubic marker represents an element. Looking at the marker in AR, you can see the element rendered inside the marker. Bringing two cubes together causes the elements to react, and the resulting compound is rendered inside each marker.
This tactile learning is really powerful. You manipulate the cubes: deciding which elements should react by picking the faces and bringing the cubes together. That's it! No on-screen manipulation at all. It's a very satisfying experience even without access to AR glasses.
- Problems with their cross approach
- Problem: cutting is annoying
- Problem: gluing is annoying
As far as I can tell, Elements 4D started as a kickstarter, where they were selling wooden cubes as a kit. At this point, I don't think these are available anymore, and they have a new approach. You print, cut and glue their AR markers together.
TODO: Why is this annoying?
Approach 1: single folding
- Problem: can't print full bleed
- Problem: the actual fold is challenging
Approach 2: legos
- Approach 3: magnetic sheets