One of the topics which is going to play a very important role in intelligent displays is Face Detection and Tracking. At the moment there are many applications in C++ and Java doing Face Detection. In the Web-Area (AS3 and Processing) it still seems a bit quiet…most likely because the power of the Languages just doesnt suffice…

In any case some interesting links regarding this topic:

http://en.wikipedia.org/wiki/Morphological_image_processing
http://www.face-rec.org/source-codes/
http://wwwiaim.ira.uka.de/computer-vision/ (german)
http://opencvlibrary.sourceforge.net/ (The openCV project is a library which provides function for real-time computer vision)
http://ivt.sourceforge.net/ (Integrating Vision Toolkit)
http://www.fujifilm.com/products/digital_cameras/f/finepix_f50fd/features/index.html (seems that FujiFilm hast integrated a very capable Face Detection technology in its newer cam models)

Why does one of them have a green box?  My Canon Camera has a feature like this, but it does not work very well from a distance, and can usually only get one or two faces.

Why does everything have to be source only?
I would love to try IVT but I can’t compile.
Preferably be compatible with vvvv or maybe a dll if it’s not too hard to interface. Why not freeframe compatible?

Run Linux and you get used to it Why can’t you compile it?

I have been thinking about Linux but I don’t have a way to set it up yet. Eventually I might try a dual boot but not for a while. Maybe Ill use a Linux box as the tracker and send it to another computer for further work.

1 I don’t know how to compile
2 I don’t have the proper tools. Nor will I have for a while. Setting up a separate box just for compiling is a little too much.
3 Way too many prereqs and no universal way to make it work. Just more bloat and dead ends.

Why not just release binaries as the developers already have tested their work. Many projects have long tutorials with hacks and hoops to get it to work.

So no way for non programmers to ever use the tools developed.

Eventually I might ‘get’ it but then it will be an ever lasting road of hoops to jump through. Hoops are still hoops regardless of how easy they become.

Nah, its not anywhere near as hard as you describe it.  You can compile on windows, I am pretty sure the GCC toolchain runs on it.  Its all free, so its just a download away.  If you follow the guide step by step you should be able to do it.

GCC? Can’t find it. Well I can find a lot but not sure what to look for.
I hope it’s as easy to get working as you say. I really want to play.

Here is a port of the GNU Toolchain to Windows, which includes GCC.

http://www.mingw.org/

See if you can get that running.  Once you have that you will need OpenCV and Qt.  If you have Touchlib working then you already have OpenCV.  To get Qt is simple, just go here and download it.  No compiling, its a simple installer.  http://trolltech.com/downloads

Now I have installed that and got it to run.
Now I can’t find out how to compile the code. Qt only takes .ui files. What am I suppose to do?

Run these in order:

That should build the binary files.

Just so you know, I found almost the same thing in one Google search, always Google three different keyword combinations before asking.

Very interesting thread ,

about Face Detection and Processing , two libraries are available and fully documented :

http://tokage.cafe24.com/facedetect/ for MAc

http://www.bryanchung.net/?page_id=251 for WINDOZ

It would be appreciated if you changed PC to Windows.  Us GNU/Linux users are out there too, and we need to be acknowledged

DOne ! ; pliz send my Apollo G to the GNU’s !

Thanks, I love the spelling too