Abstract: This paper proposes a new application for face recognition task on mobile device. The originality of the work is based on the web service, able to store the normalized distances in the database, to verify the correct building of the classifier’s model and to match the test normalized distances with the training normalized distances already stored in the database. This innovative technique is named Bridge Approach (BA). The Bridge Approach allows both to speed up the processing and to use the face recognition system in all locations where an internet network is available. The application uses the OpenCV library for dynamic targeting of face, eyes, nose and mouth and WEKA library for the classification purpose. Experimental results are carried out on 500 face images belonging to 10 individuals (50 images each). Finally, the main goal of the work is to show the performances in terms of computational time and classifier’s accuracy.
Keywords: Face Recognition System, Web Service Architecture, OpenCV, WEKA, Mobile device