ABSTRACT


These days biometric confirmation frameworks are utilized for security applications like check and recognizable proof. There are many biometric systems like fingerprint, face, and iris. Biometric system is better than pi n and password based system because they can be easily stolen or forgotten bu t i n biometric system can't easily steal and forgot because the characteristics of biometric systems are unique. J n biometric system, the fingerprint biometric system is mostly used for security purpose to identify the users according to the uniqueness of the iris fingerprint but still some problem occurs in fingerprint biometric security systems like some fake reproducible fingerprints made by silicon or other material, the fingerprints made from these materials can unlock the original fingerprint impression. I t i s viewed a one of the noticeable problem in fingerprint systems.
There is needs to develop an algorithm i n fingerprint liveness detection to identify the user either the user is genuine or fake one. So a new software based met hod Dynamic Local Ternary Pattern (DLTP) is proposed for fingerprint liveness detection with Weber's Law in this research. The Support Vector Machine (SVM) Classifier is used for fake detect ion. The results obtained from the proposed method (DLTP) texture descriptor attained optimum accuracy and clearly outperformed the reported Local Binary Pattern (LBP) and Local Ternary Pattern (LTP) texture
descriptors.