A Multi-Algorithmic Face Recognition System

The importance of utilising biometrics to establish personal authenticity and to detect impostors is growing in the present scenario of global security concern. Development of a biometric system for personal identification, which fulfills the requirements for access control of secured areas and other applications like identity validation for social welfare, crime detection, ATM access, computer security, etc. is felt to be the need of the day. Face recognition has been evolving as a convenient biometric mode for human authentication for more than last two decades. Several vendors around the world claim the successful working of their face recognition systems. However, the Face Recognition Vendor Test (FRVT) conducted by the National Institute of Standards and Technology (NISI), USA, indicates that the commercial face recognition systems do not perform up to the mark under the variations ubiquitously present in a real-life situation. Availability of a largely accepted robust face recognition system has proved elusive so far. Keeping in view the importance of indigenous development of biometric systems to cater to the requirements at BARC and elsewhere in the country, the work was started on the development of a face-based biometric authentication system. In this paper, we discuss our efforts in developing a face recognition system that functions successfully under a reasonably constrained set-up for facial image acquisition. The prototype system built in our lab finds facial match by utilizing multi-algorithmic multi-biometric technique, combining gray level statistical correlation method with Principal Component Analysis (PCA) or Discrete Cosine Transform (DCT) techniques in order to boost our system performance. After automatic detection of the face in the image and its gross scale correction, its PCA and DCT signatures are extracted. Based on a comparison of the extracted signature with the set of references, the set of top five hits are selected. Exact scale of the face is ascertained w.r.t. each of these hits by first locating the eyes employing template matching technique and then finding the inter-ocular distance. After interpolating the face to the exact scale, matching scores are computed based on gray level correlation of a number of features on the face. Final identification decision is taken amongst this set of five faces, depending on the highest score. We have tested the technique on a set of 109 images belonging to 43 subjects, both male and female. The result on this image-set indicates 89% success rate of our technique.