A Common Lisp Wrapper of OpenCV's C++ Interface - Project on hold until CLASP matures. When some of the initial bugs in CLASP are worked out I will redesign this library to use CLASP generated bindings for its C++ interface instead of CFFI.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


Welcome to my Lisp-CV repository, My name is Joe BiMedina. (Update - The creation of this project is on hold for now. It is over 2/3 complete and most bugs have been removed so you can get a lot of use/fun out of it. By the way, this has honestly been a very fun library to build and play with. So you know, I tested the speed of this library vs. Pythons OpenCv library and it is very comparable in most respects and almost all of the bugs have been ironed out in the current build. I hope to start up again soon and integrate Dr. Chris Shafmeisters CLASP(as it matures) as the backbone of Lisp-CV, instead of CFFI, hopefully making it a much faster library. The wait is designed so any possible issues in CLASP can be discovered/fixed before I begin the integration. )

Tested on Ubuntu - all versions 12.04 and up - using SBCL and CLISP and ALLEGRO CL

Warning: Work on this library is still in process. Some functions may not work as expected. I test the functions all the time so you have a pretty good shot that almost all of them would work.


Lisp-CV is a Lisp wrapper for OpenCV's C++ interface, not the C interface. The Lisp wrapper wraps around C wrappers which wrap around the OpenCV C++ functions. Due to the speediness of C and C++, the speed of the C++ functions compared to the C wrappers for them is almost identical.

This install tutorial is made for people wanting to help work on the project. Thus no build file is included so you'll have to install everything manually. Don't worry it's easy. This is intended so you can familiarize yourself with how everything is put together. As stated previously, Lisp-CV is known to work on Ubuntu versions 12.04 through 14.04. It should work on any system that OpenCV, SBCL, CLisp, Allegro Lisp, CFFI, ASDF and QuickLisp can be compiled on. There is nothing but Lisp, and C++, C and Python in the whole project.

If you have gotten Lisp-CV to work on another system and would like to add the install instructions to this tutorial just send me an e-mail at wnetai@gmail.com with the install instructions included. They will be added to this file and you will be given credit for your addition unless you decline.

If you would like to contribute a function to this project, you will need to first find out if the C wrapper for it is available. Don't worry 550+ C bindings are available and they can be found in the below files.(See the INSTALLATION section for the location of these files.)


If you run into a Lisp function that doesn't work suddenly after compiling the latest .cpp files, and it worked before, a change in the C bindings might be the reason. Check the above .cpp files to make sure the C bindings weren't changed in anyway. Also, I make a lot of changes every day, so a non-working function could be the result of something I missed. I am very good at discovering these errors so it shouldn't be long before I fix said issue. You can hasten the fix by sending a bug report.

I write a code example and documentation for each function I add to the library, so it is very well documented. In the lisp-cv-master/examples/examples.lisp file there are over 12,000 lines of code examples and documentation. If you would like to contribute a Lisp function, I ask that you also write an example of how to use it. It doesn't have to be a large example it only has to show what the function does and give an idea how it is used in the most basic way. See lisp-cv-master/examples/examples.lisp for an example of the examples I already included.

To add contributions just Fork my repository at:

https://github.com/W-Net-AI/LISP-CV and send me a Pull Request.

or send them to me in an e-mail at wnetai@gmail.com.


To use Lisp-CV you will need to first install the development build of OpenCV here https://github.com/Itseez/opencv and GitHub member arjuncomar's fork of opencv_contrib here https://github.com/arjuncomar/opencv_contrib. Instructions to build both are in the Readme.md at the former link. You will also need Lisp environment, of course.

Note #1: I will make a Makefile or script to install Lisp-CV as library progresses but for now this install tutorial is made people who want to work on the project. So all steps to install must be done by hand so you get familiar how the library works...don't worry though it's easy. If you do end up helping on the project you might want to make a script to speed up some of the process.

Clone sources from official repository (https://github.com/W-Net-AI/LISP-CV) first and them extract them to location reachable by ASDF. Discussing ways to make ASDF look for systems definitions in custom directories is out of scope of this instruction, please refer to ASDF manual.

Note: You may want to change the name of the extracted LISP-CV source folder from "LISP-CV-master" to "lisp-cv-master", to prevent any possible unforeseen issues. I don't think there would be, but, better to be safe. The caps were just used to make the LISP-CV repo look nicer and do not appear anywhere in the source code.

If you built OpenCV with the extra modules correctly per this link:


You should have these files present on your machine:

"Note: for the time being, until an error in the opencv_contrib module is ironed
out, the opencv_generated.cpp and opencv_generated.hpp will be provided in the: 



directories. You will still need to build OpenCv with the instructions at the above 
link, but when your done, will need to place these files at the appropriate paths below. 
This should work fine but, due to the error, I am unable to test the install at this time 
to verify it."

You will need thes files as well. They can be found in the Lisp-CV src and include directories:


Use any C++ compiler to compile all the above files. In example provided I will use GCC:

g++ -Wall -shared -fPIC -o <filename>.so <filename>.cpp

Then install resulting so-files manually to a location on your PATH(i.e usr/local/lib in Linux).

Now you can start your Lisp and load LISP-CV:

(asdf:load-system :lisp-cv) ;The main package.

(asdf:load-system :gc) ;Automatically finalized functions.

You're all done! Now you can switch your package to LISP-CV's for extra comfort:

(in-package #:lisp-cv)

(in-package #:gc) 

Let's do some tests just to be sure. You'll need to compile this...

(defun imread-example-2 (filename)
  "Open the image FILENAME with IMREAD and show it in a window."
  (let ((window-name "IMREAD Example 2"))
    (with-named-window (window-name +window-normal+)
      (move-window window-name 759 175)
      (with-mat ((image (imread filename 1)))
        (if (empty image)
          (return-from imread-example-2 (format t "Image not loaded")))
        (imshow window-name image)
          (let ((c (wait-key 33)))
            (when (= c 27)

And then evaluate this at the REPL...

(imread-example-2  <path-to-any-image>)

If all went as planned, you should see your picture in a window!


You can learn to use this library by looking at the examples in:


If you want to know if a function has been wrapped and is available, do a search for the C++ function name in the examples.lisp folder. All functions are documented and all have the C++ function declarations above the Lisp one. Learning how to search this file will help answer a lot of questions about this Library and put incredible amounts of code snippets right at your fingertips. Every function provided in this library has an example program associated with it and all of the names of the examples are formatted as -example.

You can also learn how to use the library by comparing the OpenCV C++ tutorials at this link:


to the VIDEO-CAPTURE-EXAMPLE in examples.lisp.

By comparing the tutorials at this link:



and by comparing the tutorial at this link:


to the SET-MOUSE-CALLBACK-EXAMPLE. Comparing the C++ code to the Lisp code will give you an idea of how to convert C++ OpenCv code you find online to Lisp.

All the constants are defined as so:

If a constant in OpenCV's C++ interface is WINDOW_AUTOSIZE

The Lisp version is +window-autosize+

If a constant in OpenCV's C++ interface is CV_8UC3, it's Lisp version is +8UC3+

I always add the plus signs to both ends and I always change the underscores to dashes and I always remove the "CV_".


Feel free to e-mail at me wnetai@gmail.com for any of the following reasons:

Bug reports

If you would like to help develop this project(I intend to make it complete)

Request for features

If you have gotten Lisp-CV to work on a system other than specified at the top of this page (Remember to include how you got it done!)

If you have a function and example that you would like to add

If you would like add a code sample to the examples folder(something you made with Lisp-CV)

If you would like to add some other unspecified addition to the library