The process of verifying a person’s identity, also called authentication, plays an important role in various areas of everyday life. Any situation with user interaction where the identity is required needs a means to verify the claimed identity. One of the more obvious and commonly known application areas for identity verifying technologies, i.e. authentication, is the Logical Access Control to computer systems, where authenticity is normally established by confirming aclaimed identity with a secret password or PIN code.Traditional methods of confirming the identity of an unknown person rely either upon some secret knowledge (such as a PIN or password) or upon an object the person possesses (such as a key or card). But testing for secret knowledge or the possession of special objects can only confirm the knowledge or presence, and not, that the rightful owner is present. In fact, both could be stolen. Conversely, biometric technology is capable of establishing a much closer relationship between the user’s identity and a particular body, through its unique features or behavior.
Biometric verification performs comparison of biometric template with the one it has on records. Face recognition is one of the techniques used in biometric verification. While performing face recognition on mobile platform it does not only suffer
from the same problems of a computer based system, such as illumination, occlusion and pose variations,
but is also limited by other factors: Limited Processing Power, Limited Memory . To implement face recognition based authentication for incoming calls on mobile, existing algorithms suffers from recognition time and Accuracy Tradeoff i.e. increasing robustness will increase the time of recognition.
To implement Real Time Recognition i.e. recognition must be performed within few seconds to support incoming call authentication and to make application robust under different illumination conditions. We use Retina illumination normalization  followed by simple, but robust and efficient face recognition algorithm for implementation in mobile devices. It is based on Local Binary Patterns (LBP) features , which have shown to be a powerful and computationally efficient feature extractor. Then Chi-square based classification method is used for Authentication.
The algorithm constitutes engine of a new face authentication for incoming call application. It restrictsuser to receive call unless proposed entity’s face is not verified with registered entity’s face image. The implementation is made for the Google Android platform, using OpenCV libraries for image processing. It is important to take into account that all the processing is done in the mobile domain, without the need of any external computation.
Figure 1.Flow Chart for Authentication on Incoming Call
An authentication method based on face recognition is designed for restricting access of incoming call for un-authorized entity. It will take...