Face recognition using haar classifier:How to train and compare with stored images
Hi friends
I'm gonna create a simple face recognizer.
In fact I can detect the face using haar classifier, but I want to compare it with a stored image in my file in order to determine that the detected face is who? (as an example tell me the owner name of the detected image)
Can someone help me?
I read some source codes in the different websites but unfortunately they are not clear for me because I'm beginner.
I greatly appreciate if you guide me step by step or put the source code in answer section.(c++ code)
haar classifiers are meant to do : "face or not" (or , whatever object they were trained for). it's a binary classifier, you want multi-class.
darn that dream, wrong tool for the job. look at opencv's face-recognition or even here for more state-of-the-art ideas.
again, there's no such thing as a 'simple' solution.
@berak is there any sample code?
@berak: I got it.I should use fisher face but could you please give me link which I can use?
you will need the contrib repo
samples are here
If you can be patient till the end of the month, the OpenCV Blueprints will have a chapter with accompagnied code on face recognition interfaces in OpenCV, as part of a biometrics project. But they big tutorial with the face recognition module tells you alot also.