From 696b756e903c320dc630a9e897b3e74ba18b83dc Mon Sep 17 00:00:00 2001 From: Antoine Villeret Date: Tue, 8 Oct 2013 10:00:10 +0000 Subject: cleanup build system does not work yet on windows, but working on it... svn path=/trunk/externals/pix_opencv/; revision=17215 --- README.txt | 91 ++++++++++++++++++++++++++++++++++---------------------------- 1 file changed, 50 insertions(+), 41 deletions(-) (limited to 'README.txt') diff --git a/README.txt b/README.txt index e056533..ed941f3 100644 --- a/README.txt +++ b/README.txt @@ -1,42 +1,51 @@ -updated for Linux on 03/01/2013 - -get the source version through svn : - -svn co https://pure-data.svn.sourceforge.net/svnroot/pure-data pure-data -cd pure-data/trunk/externals/pix_opencv - -you need a working build toolchain - -pix_opencv depends on OpenCV 2.x and >1.5 (for old externals) - -===== GNU/Linux ===== - -# (actually only tested in GNU/Linux Ubuntu) -# this new version requiered OpenCV 2.3, available with apt-get since 12.04 : -apt-get install libopencv-dev - -# for other systems, you can build the lib from source, see this : -# http://opencv.willowgarage.com/wiki/InstallGuide - -# get the source : -svn co https://pure-data.svn.sourceforge.net/svnroot/pure-data/trunk/externals/pix_opencv pix_opencv - -cd pix_opencv -make - -# To use the binaries you can either put them into the same folder as your patch or to add this folder to the search paths of Pd - -===== MAC OSX (intel and powerPC) ===== - -# install opencv with fink : -fink install opencv-dev - -# if you don't have fink, you can build it from source by following this guide : -# http://opencv.willowgarage.com/wiki/InstallGuide - -# then get the sources and compile : -svn co https://pure-data.svn.sourceforge.net/svnroot/pure-data/trunk/externals/pix_opencv pix_opencv +1. Prepare you machine +You need to have puredata and Gem installed on your system. +And you need to know where they are… + +You also need a compiling toolchain, e.g. GCC on Linux/Mac or Microsoft Visual C++ on Windows. + +2. Get the sources : +Download the pix_opencv sources from Puredata external repository : +svn co svn://svn.code.sf.net/p/pure-data/svn/trunk/externals/pix_opencv +or if you don't have svnl, can get a "snapshot",i.e. a zip file with the current code, here : +https://sourceforge.net/p/pure-data/svn/HEAD/tree/trunk/externals/pix_opencv/ + +3. Get opencv : +Ubuntu/Linux +The easiest way to get it is to install it from the repository : +sudo apt-get install libopencv-* + +Mac OSX : +The easiest way is to get from some packaging repository like Macports, fink or homebrew. +with fink : +sudo fink install opencv-dev +with macport : +sudo port install opencv + +Windows : +Download the binary release on opencv.org +You can follow the quickstart guide on opencv.org to setup environment variables. + +4. Build the Sources : + +Ubuntu/Linux & MacOSX +Go to the fresh created folder, i.e. « pix_opencv » and build the things : cd pix_opencv -make PD_PATH=/path/to/pdextended-app/Contents/Resources/ - -# PD_PATH should be ended by something like that : Pd-0.43.4-extended-20121231.app/Contents/Resources/ +make + +If you see something like « can’t find m_pd.h », you need to tweak pd’s path with this command : +make --with-pd=/path/to/pd +If you see something like « can’t find Base/GemPixObj.h », you need to tweak Gem’s path with this command : +make --with-gem=/path/to/gem +You can combine options like : +make --with-pd=/path/to/pd --with-gem=/path/to/gem + +Windows +You need Visual C++ (at least the free express edition) to build pix_opencv on Windows. +There is a solution in build/vs2010 folder. +It was made with Visual C++ Express 2010 and surely need some tweaks before producing any DLL. +You need to change the include and the library paths. +The include must contain the following : + - puredata include (pd-extended) or src (vanilla) folder : path\to\pd\src or path\to\pd-extended\include + - Gem include folder : path\to\Gem + - OpenCV include path, wich is ususally $(OPENCV_DIR)\..\..\include if you setup OPENCV_DIR according to Windows quickstart guide on opencv.org. -- cgit v1.2.1