The era of the Kinect motion sensor for Microsoft Xbox gaming consoles has passed. If you still have one and wondering how to use it (except for gaming), you might want to try 3D scanning. In this article, I’ll share my experience with the Xbox 360 sensor in a few steps you can follow.
Simply put, Kinect is a smart camera (well, if you like more complicated explanations check this) that allows you to control your Xbox gaming console using gestures. Only two generations have seen the light of day – the Xbox 360 and Xbox One (along with Windows versions for developers), with Kinect development being discontinued in 2017 due to a lack of popularity and changes in business strategy.
Plug it in
An additional adapter is required for connection to a PC.
Next, you must install the system drivers. Although the latest version is 2.0, I recommend choosing 1.8 on Windows 10 – they work without any problems, unlike the newer version (according to reports from others, I was not the only one who had problems with version 2.0).
In addition to the drivers, you can look for the Developer Toolkit, which contains a number of interesting demonstrations of Kinect’s capabilities, as well as tools and tutorials for developers.
The very purpose of the Kinect is not high-quality 3D scanning and therefore it has some limitations. Even so, with a little effort, it is possible to create interesting things. You can’t make a model without scanning software. I’ve tried two of them, ReconstructMe and probably widely known Skanect.
Scanning can typically be done in two ways. The first is to scan an object from all angles with a camera in hand, which is suitable for most situations.
The second is to rotate the scanned object in front of a static camera. If your goal is to create a bust (mine was), you’re practically taking a selfie in which you have to stand in front of the sensor and rotate 360 degrees (the rotation speed depends on the desired quality). Sounds easy, until you do it for the 20th time. 😀
Simple scanning software that is also completely free for non-commercial use. Just set the correct sensor input type and mode (in my case it was a selfie, designed specifically for busts).
Tutorials on how to use the software (including videos) can be found directly on the website. Export can be done to STL, OBJ, 3DS, and PLY. The .stl format is most commonly used for 3D printers, .obj for editors like Blender, .3ds for Autodesk 3ds Max, and finally .ply, which is a general format for saving data from 3D scanners.
Summary: the software configuration is easy – most of the things are done automatically in the background (filling gaps, colorization). It’s harder to work with the light transition. I had a problem with exporting to .obj. However, for hobby use it is ideal.
Compared to ReconstructMe, it is a more robust software. The free version has some limitations which are not practically noticeable in the case of personal use. Except for the already mentioned export formats, it also offers VRML and direct upload to Sketchfab and Shapeways (Pro version).
Summary: requires more playing with configuration before scanning, some steps need to be done additionally afterward (previously mentioned filling gaps, colorization, etc.). This software is therefore suitable for more advanced users.
Thanks for reading! If you have your own experiences in this area, feel free to share them in the comments.
This article is an edited and updated version of my 2021 article published on the website of my home scout troop, Patnáctka.