Pattern recognition and similarities measure

I'm working on my first project in pattern recognition field, employing Yale image's database. I'm using log-gabor transformation on the images then applying a developed mask in order to select the effective features. But I don't know what to do to classify the results as separate patterns and measure the similarities between the test image and the patterns. I'll be very grateful for guiding me to the steps.

Answer 1

[Image Analyst]

Are you trying to find similar images, like they do in CBIR (show me all photos of beaches in this stock photography database"), or are you trying to find a particular shape or pattern like they do in target detection/acquisition/identification ("Find the tank in this image")>

[hend]

I try to identify an unknown face.

[Image Analyst]

OK, totally different. There is no face recognition toolbox so you're on your own, or, actually, you can look up the vast database of papers to find an algorithm you like: http://iris.usc.edu/Vision-Notes/bibliography/contentspeople.html#Face%20Recognition,%20Detection,%20Tracking,%20Gesture%20Recognition,%20Fingerprints,%20Biometrics

Pick one you like, then program it up.