LiveScan3D is a system designed for real time 3D reconstruction using multiple Kinect v2 depth sensors simultaneously at real time speed. There is also a new, experimental, branch that works with Azure Kinect, link here, updates about that branch will be posted here.
The produced 3D reconstruction is in the form of a coloured point cloud, with points from all of the Kinects placed in the same coordinate system. Possible use scenarios of the system include:
- capturing an object’s 3D structure from multiple viewpoints simultaneously,
- capturing a “panoramic” 3D structure of a scene (extending the field of view of one sensor by using many),
- streaming the reconstructed point cloud to a remote location,
- increasing the density of a point cloud captured by a single sensor, by having multiple sensors capture the same scene.
You will also find a short presentation of LiveScan3D in the video below (click to go to YouTube):
At the moment connecting multiple Kinect v2 devices to a single computer is difficult and only possible under Linux. Also, the number of the devices or capture speed might be low, because of the limitations of the PCI-E bus.
Because of those limitations, in our system each Kinect v2 sensor is connected to a separate computer. Each of those computers is connected to a server which governs all of the sensors. The server allows the user to perform calibration, filtering, synchronized frame capture, and to visualize the acquired point cloud live.
To start working with our software you will need a Windows machine with the Kinect for Windows SDK 2.0, and at least a single Kinect v2 device. You can either build LiveScan3D from source which, or you can download the binary release. The current version has to be compiled using Visual Studio 2015 for compatibility with the OpenCV dll versions that are included.
Both the binary and source distributions contain a manual (in the docs directory) which contains the steps necessary to start working with our software (it won't take more than a couple of minutes to set up).
If you have any problems feel free to contact us: Marek Kowalski [email protected], Jacek Naruniec [email protected]. We usually answer emails quickly (our timezone is CET).
For details regarding the methods used in LiveScan3D you can take a look at our article: "LiveScan3D: A Fast and Inexpensive 3D Data Acquisition System for Multiple Kinect v2 Sensors" once it is released.
While all of our code is licensed under the MIT license, the 3rd party libraries have different licenses:
- nanoflann - https://github.com/jlblancoc/nanoflann, BSD license
- OpenCV - https://github.com/Itseez/opencv, 3-clause BSD license
- OpenTK - https://github.com/opentk/opentk, MIT/X11 license
- ZSTD - https://github.com/facebook/zstd, BSD license
- SocketCS - http://www.adp-gmbh.ch/win/misc/sockets.html
If you use this software in your research, then please use the following citation:
Kowalski, M.; Naruniec, J.; Daniluk, M.: "LiveScan3D: A Fast and Inexpensive 3D Data Acquisition System for Multiple Kinect v2 Sensors". in 3D Vision (3DV), 2015 International Conference on, Lyon, France, 2015
- Marek Kowalski [email protected], homepage: http://home.elka.pw.edu.pl/~mkowals6/
- Jacek Naruniec [email protected], homepage: http://home.elka.pw.edu.pl/~jnarunie/