This is an OpenCV-based motion sensing game controller. It is capable to be deployed on Raspberry Pi or Linux PC.
If you want to deploy it on Linux PC, make sure,
- gcc supports C++11.
- OpenCV 3.x installed.
- Built-in camera or USB camera connected.
If you want to deploy it on Raspberry Pi
- gcc supports C++11
- OpenCV 3.x installed.
- raspicam library installed https://github.com/cedricve/raspicam.git
- Camera connected via CSI port
Build on PC: make -j4 Build on Raspbarry Pi : make RPI=1 -j4
Usage: Under ./bin directory,
-
Change the IP address and port number in connection_config to server IP and port number you want to connect to. This is called "connection configuration" file.
-
Caliberate reference histogram ./caliberate [histogram data file name] Cover all the rectangles with your palm, follow the instructions on window title. Caliberated histogram data will be stored in [histogram data file name], to be used later.
-
Make sure server is alreading waiting to accept. A demo of the server https://github.com/OverflowException/baymaxmatch.git Then run the controller ./controller [connection configuration file name] [histogram data file name]
note: When running controller, make sure your hands and face are the only skin-colored objects in sight. Especially, when naked, or with sleeves rolled up, this controller is completely unusable.