Fuzzy rule based approach for face and facial feature extraction in biometric authentication

Human biometrics are regarded as groundbreaking tools for preserving privacy and security in many computer vision applications such as, biometric authentication, secure access control, visa processing, and border checking. Biometric features such as, fingerprint, face, retina, iris patterns, voice waves, palm print and signatures are commonly used in biometric authentication. However, in recent years face and facial features have become as the key attributes in biometric authentication due to their uniqueness and robustness. This manuscript proposes a robust and efficient scheme for extracting facial biometric features from the face images using fuzzy rules. The proposed system estimates face and facial features by analyzing color components in the human face images. In this paper we have used a combined color space consisting of HSV and YCbCr models to extract the color components. The face skeleton is determined based on the largest connected skin color area. Facial features such as, eyes, nose, mouth are then localized from the face skeleton with the knowledge of the face geometry. The algorithm is characterized by its simplicity and inexpensive computational requirements. Experimental evaluation confirms that our proposed system provides better performance comparable to other methods.