Kinect Mouse Tracker
Processes live Microsoft kinect data to deduce a mouse's position inside an enclosed pen. Made during my internship at NERF.
Git repository
In september 2018 I got the opportunity to do a 4-week internship with Karl Farrow's lab at NERF (Neuro-Electronics Research Flanders). The main focus of Farrow Lab is "Dissecting the lines of communication connecting the eye with behaviour". To achieve this the lab uses several behavioural set-ups. Among these is the one Anna Chrzanowska uses in her PhD research, with monitors on multiple sides of an open pen. See the first prototype I helped build below.
To collect experimental data and/or change visual stimulation patterns in real-time it is necessary to know the mouse's position in the pen. For the 2D postion the ir depth sensor of a Microsoft Kinect is used. Processing this data is where the application "Kinect Mouse Tracker" I developed comes in. Written in C++, it uses the computer vision library OpenCV to determine and record the mouse's postion with a delay of a little over 50ms.
