Guile-CV - Image Processing and Analysis in Guile - is a Computer Vision functional programming library for the Guile Scheme language.

Guile-CV is based on Vigra (Vision with Generic Algorithms), enhanced with additional algorithms (Image Textures, Delineate, Reconstruction and many more), all accessible through a nice, clean and easy to use high level API.

Guile-CV is natively multi-threaded, and takes advantage of multiple cores, using high-level and fine grained application-level parallelism constructs available in Guile, based on its support to POSIX threads.

Latest News

November 2018

Guile-CV version 0.2.1 is released. See the News page for the list of visible changes since the last release, as well as older news. See Guile-CV git summary and Guile-CV git log for a complete description.


Guile-CV objective is to be a robust, reliable and fast - Image Processing and Analysis - Computer Vision functional programming library for the Guile Scheme language. Guile-CV also wants to be easy to use, study, modify and extend.

Guile-CV can be used as an educational sofware, a research toolbox but it can also be used 'in production': Guile-CV is robust, reliable and fast, and we will make sure Guile-CV remains robust, reliable and fast as it grows.


Documentation for Guile-CV is available online, as is documentation for most GNU software. You may also look at the Guile-CV manual by running info guile-cv or find it, if you installed Guile-CV manually and did run make install-html, by looking at $prefix/share/doc/guile-cv/guile-cv.html/. If you installed Guile-CV using a distribution package, you may find it, perhaps, by looking at /usr/share/doc/guile-cv/ or similar directories on your system


Guile-CV also has a Savannah project page.


Guile-CV is a free software, distributed under the terms of the GNU General Public License GPLv3 or higher. You must be aware there is no warranty whatsoever for Guile-CV. This is described in full in the license.