What Are the Best Tips for Making a DIY 3D Scanner?

N. Kalu

Building a DIY 3D scanner is a very ambitious project for most enthusiasts. To make a scanner, the proper design must first be chosen, and the appropriate parts obtained. A 3D scanner is a machine capable of producing three dimensional computer models from real world objects, which can then be used to create computer generated images or loaded into a 3D printer to produce a three dimensional copy of the original. Scanning devices or webcams are often used to capture the image of the object. As 3D scanners are not widely commercially available, most individuals who want one have to create a DIY 3D scanner.

3D scanners can load images into 3D printers in order to print a three-dimensional object.
3D scanners can load images into 3D printers in order to print a three-dimensional object.

There are several designs that can be used to make a DIY 3D scanner. The laser scanner uses a laser that can project a line of light against a flat surface. Ordinary lasers produce a beam, so the line must be created using a special lens or an oscillator. By pairing a webcam with a line of laser light, a computer can use triangulation to determine the shape of the object the laser reflects off.

Another kind of DIY 3D scanner uses two webcams, or, sometimes, specialized webcams with two lenses and image sensors. These systems mimic human depth perception and calculate the shape of the object based on the difference in image between the two webcams, known as the "parallax." This type of scan requires only a single image, so cheap digital cameras may be used instead of webcams. Most builders prefer to use webcams anyway because of their low cost, good software support, wide availability, and video capture option.

One type of more advanced DIY 3D scanner is a structured light scanner. This scanner uses one camera or webcam and a projector. The projector projects a complex digital reference image similar to a 2D barcode, except full color. The computer then attempts to triangulate every visible point of the target object using the known positions of the camera and the projector, as well as the known reference image. Although this method is very fast, it can sometimes fail when scanning objects that have different shades or colors, because the computer can fail to distinguish the target object from the reference image.

These designs all capture only one side of an object. While this is fine for users who are only creating a static image or a mask, it is a severe limitation for those who wish to scan objects for placement in virtual environments or video games. To solve this problem, any of these systems can be paired with a turn table to rotate a small object while the scanner is active. The rate of rotation must be known for the scanning software to work. Many designs use old record players or pottery wheels as turn tables.

You might also Like

Readers Also Love

Discussion Comments


@Iluviaporos - It makes me think of those races that are sponsored by the army, where contestants have to design a machine capable of crossing rough terrain without guidance and return with a "wounded soldier" or at least the equivalent weight.

I read once that one of the biggest problems was coming up with a way of allowing the machine to scan the terrain ahead of itself and calculate the best trajectory.


@umbra21 - It probably depends on how much you want to commit to making it from scratch. I imagine there are different components to consider, like the software and the recording equipment and everything, that could be bought separately and it's just a matter of hooking them together.

But if you have a specific end in mind, or you want to cut costs, you'd have to do particular parts from scratch, which would require a lot of knowledge, or at least the ability to follow specific instructions. Still, you might only want to build the software aspect to your own specifications and be able to just hook it up to an existing video camera or something like that.

It really depends on your goals and your capability.


I wonder if a scanner is actually even more ambitious than making a 3D printer. It seems like it would be more difficult to figure out how to feed real life information into a computer to turn it digital then to program a computer to move in a way that will create something new.

I mean, I've even seen DIY projects online for making 3D printers out of Lego bricks, so they can't be that complicated once you get past the need for some way to translate the design.

Post your comments
Forgot password?