# SwisTrack
> Mediated Wiki article. Canonical URL: https://mediated.wiki/source/SwisTrack
> Markdown URL: https://mediated.wiki/source/SwisTrack.md
> Source: https://en.wikipedia.org/wiki/SwisTrack
> Source revision: 1354161227
> License: Creative Commons Attribution-ShareAlike 4.0 International (https://creativecommons.org/licenses/by-sa/4.0/)
{{Infobox software
|logo = Image:Swistracklogo.PNG
|caption =
|screenshot =
|developer = [Nikolaus Correll](/source/Nikolaus_Correll), Pierre Roduit, Thomas Lochmatter and others
|latest_release_version = 4.0.0
|latest_release_date = {{Start date and age|2008|02}}
|operating_system = [Linux](/source/Linux), [Windows](/source/Microsoft_Windows), [Mac OS](/source/Mac_OS)
|genre = [Video tracking](/source/Video_tracking)
|license = [Adaptive Public License](/source/Adaptive_Public_License) ([Open Source](/source/Open-source_license))
|website = {{URL|swistrack.sourceforge.net}}
|programming language = [C++](/source/C%2B%2B)
}}
'''SwisTrack''' is an [open-source](/source/open-source) tool for tracking [robot](/source/robot)s and other objects using a [camera](/source/camera) or a recorded [video](/source/video) as input source. It has mainly been developed by Distributed Intelligent Systems and Algorithms Laboratory (DISAL) and the LPM Vision Group at [EPFL](/source/EPFL) in Lausanne, Switzerland.[{{cite book |chapter-url=https://infoscience.epfl.ch/record/125704 |title=2008 IEEE/RSJ International Conference on Intelligent Robots and Systems |first1=Thomas |last1=Lochmatter |first2=Pierre |last2=Roduit |first3=Chris |last3=Cianci |first4=Nikolaus |last4=Correll |first5=Jacques |last5=Jacot |first6=Alcherio |last6=Martinoli |chapter=SwisTrack - A Flexible Open Source Tracking Software for Multi-Agent Systems |date=2008 |pages=4004–4010 |doi=10.1109/IROS.2008.4650937 |isbn=978-1-4244-2057-5 |s2cid=2527599 |access-date=January 20, 2024}}]
== Architecture ==
SwisTrack is written in [C++](/source/C%2B%2B) using the [wxWidgets](/source/wxWidgets) library for [GUI](/source/GUI) and networking. It uses [Intel](/source/Intel)'s [OpenCV](/source/OpenCV) library for fast [image processing](/source/image_processing) and contains interfaces for [USB](/source/Universal_Serial_Bus), [FireWire](/source/FireWire) and [GigE](/source/Gigabit_Ethernet) cameras, as well as [AVI](/source/Audio_Video_Interleave) files.[{{cite book |chapter-url=https://infoscience.epfl.ch/record/85929 |title=2006 IEEE/RSJ International Conference on Intelligent Robots and Systems |first1=Nikolaus |last1=Correll |first2=Gregory |last2=Sempo |first3=Yuri Lopez |last3=De Meneses |first4=Jose |last4=Halloy |first5=Jean-Louis |last5=Deneubourg |first6=Alcherio |last6=Martinoli |chapter=SwisTrack: A Tracking Tool for Multi-Unit Robotic and Biological Systems |date=2006 |pages=2185–2191 |doi=10.1109/IROS.2006.282558|isbn=1-4244-0258-1 }}]
Since version 4, SwisTrack's pipeline is built by the user from components which interact with each other and pass data through data channels. There are six data channels: input, [grayscale image](/source/grayscale_image), color image, binary image, particles and tracks. Individual components work only with some of the data channels. Each data channel that the component works with can be in one of the three modes: read, write and edit (i.e. both read and write).