Multimodal Approach with Face and Iris to enhance the efficiency of Biometric System Using Soft Biometrics
Biometric technology is the science of identifying human being by extracting a feature set from data and comparing with template store in database. A biometric system is used for identifying the person either genuine or imposter by using their physiological traits (hand geometry, face, fingerprint, Iris etc) and behavioral traits (voice, gait, signature etc.). A Biometric system which uses only single trait for recognition is known as unimodal biometric system. The performance of unimodal system is not good due to some limitation such as noisy sensor data, non-universality etc. Unimodal biometric system uses only single trait of biometric for recognition and do not provide better authentication for highly secured application. Multimodal biometric systems solve all the limitations related to unimodal biometric system. Multimodal combines different physical and behavioral traits such as face and finger print, fingerprint and signature etc. A new approach multimodal was developed to overcome the limitation of unimodal system and also improve the security. There are also some other traits such as skin color, age, height, hair color, eye color, gender called “soft biometric traits”. Soft biometric traits do not provide reliable authentication because the nature of these traits are not permanent. Due to the lack of permanence and distinct property in soft biometrics, it can be used with other traits for improving performance of biometric system. This paper proposes a framework by combining physical traits (face and iris) with soft biometric trait (eye color) for enhancing biometric system security and performance.