GoCam

GoCam is an open source (GPL) project for developing a computer vision system that analyses video capture of a go game and extracts the moves made by the players. It is implemented in C++ and uses the excellent CImg Library for image processing. Currently, the system is able to find the playing grid from still images.

System requirements

Operating system: The system has been developed on GNU/Linux and Mac OS X systems, but it should work on any decent unix system. It might compile on Windows too - at least the CImg Library also promises to support Windows.
Libraries: The template-based CImg Library is included in the source, but it requires the ImageMagick library to load compressed image formats, such as jpeg.

Download

gocam-0.3.tar.gz - C++ implementation.
gocam-0.1.tar.gz - The first prototype version implemented in MATLAB.
The ImageMagick library and tools are included in many GNU/Linux distributions. For Mac OS X, the ImageMagick can be obtained from the Fink project. The main page of the library is http://www.imagemagick.org/.

Documentation

The general ideas and algorithms behind GoCam are described in the report of the first prototype version implemented in MATLAB: gocam.ps.gz (4.1M), gocam.pdf (504k)

The C++ source code is documented with Doxygen. See the html directory in the source package.

Author

Teemu Hirsimäki. See my home page for more information. I have not done much on the project since summer 2005, but feel free to contact me if you are interested in the project. My email address is thirsima at domain iki.fi.