How Your Phone Sees a Picture and Builds a 3D World
Point a tablet's camera at a square black-and-white pattern on a piece of paper, and suddenly a 3D dragon pops up on the screen, standing right on top of the paper. It looks like magic, but it is really math: the camera is measuring the corners of that pattern about 30 times every second to figure out exactly where it is in space.
What You'll Learn
- What a 'marker' is in marker-based AR, and why it has to be black and white with sharp corners - How a camera turns a flat picture into a 3D position (distance, tilt, and rotation) - The difference between marker-based AR and markerless AR (the kind that uses the floor or walls instead) - How to build and test your own paper marker
The Marker: A Picture a Computer Can Measure
A marker is usually a square with a bold black border and a unique pattern inside, like a simplified QR code. Apps like Vuforia and ARToolKit were built around this idea back in the early 2000s. The square shape matters: a computer program can find the four corners of a square from almost any angle, and once it has four corners, it can calculate exactly how far away the marker is and how it's tilted.
From Flat Corners to 3D Position
Here is the clever part. If the app already knows the marker is a perfect 6 cm by 6 cm square, it can compare that to how the square LOOKS on camera. If the square looks small and squished into a thin parallelogram, the phone is far away and looking at it from an angle. If the square fills the screen and looks perfectly square, the phone is close and facing it directly. This math is called 'pose estimation,' and it runs continuously, which is why the 3D dragon stays glued to the paper even as you move the camera around it.
Marker-Based vs. Markerless AR
Marker-based AR (what we just described) needs a printed image to work. Markerless AR, used in apps like Pokemon GO and IKEA Place, instead scans the room for flat surfaces and textures using a technique called SLAM (Simultaneous Localization and Mapping). Marker-based AR is simpler and more reliable in classrooms because the marker never moves, but markerless AR feels more magical because objects can appear anywhere, not just on a special piece of paper.
Software specifically hunts for sharp 90-degree corners because they are easy to detect even in blurry or poorly lit video. A fuzzy circle or a photo of a cat is much harder for a computer to lock onto reliably than four crisp corners.
Match each AR term to what it actually does.
Terms
Definitions
Drag terms onto their definitions, or click a term then click a definition to match.
Why do most AR markers use a black square with sharp corners instead of a colorful photo?
If a marker looks small and squished into a thin shape on camera, what does that tell the app?
Build and Test a Paper Marker
1) Draw or print a 6 cm black square outline on white paper with a unique smaller pattern inside (stripes, letters, or shapes). 2) Using a free AR marker app (like an ARToolKit demo or a classroom AR app your teacher provides), scan your marker. 3) Slowly move the camera closer, farther, and at an angle while recording what happens to the 3D object on screen. 4) Write down 3 observations about how the object's size and tilt changed with the camera's position, and turn in your marker plus your notes.
Flashcards — click each card to reveal the answer
Want to keep learning?
Sign up for free to access the full curriculum — all subjects, all ages.
Start Learning Free