Introduction

ofxFaceRecognizer an addon for openframeworks that implements the face recognition tutorial from the opencv site: http://docs.opencv.org/modules/contrib/doc/facerec/facerec_tutorial.html

It needs opencv 2 or higher.

You extract a face from a video or a static image and feed it in to the recognizer.

At the start the recognizer either loads an already exciting .yml file that contains the recognition model or it creates a new model from the images it finds in the faces folder.

You can select between 3 different recognition methods: eigenfaces, fisherfaces or LBPHFaces.

I am supplying some example faces created by the Tarrlab at Brown University. http://wiki.cnbc.cmu.edu/Face_Place

You notice that the image file names (MF0907_1100_HA.tiff) contain unique IDs to label each person. Each person has more then one image but always the same ID. This ID is used by ofxFaceRecognizer to label each person separately.

I used ofxCv align eyes and crop functions to prepare the images before placing them in the faces folder.

Installation

Download the source from GitHub, unzip the folder, rename it from ofxFaceRecognizer-master to ofxFaceRecognizer and place it in your openFrameworks/addons folder.

Or use project generator to build project and then:

modify Project.xcconfig as described in ofxOpenCV2461

add to project-> build settings -> other linker flags ../../../addons/ofxOpenCv/libs/opencv/lib/osx/opencv.a

add to targets-> build settings ->library search paths $(OPENCV_LIBS_PATH)

Dependencies

ofxOpenCV2461 https://github.com/pkmital/ofxOpenCV2461 Requires your project to link against ZLib and OpenCL.

ofxCv https://github.com/kylemcdonald/ofxCv

ofxYAML https://github.com/satoruhiga/ofxYAML

Operating systems

ofxFaceRecognizer has only been tested on osx 10.10 with OF 0.8.4

Images

This addon has been used for "Level of Confidence": https://github.com/antimodular/Level-of-Confidence

Screen shot: