HYBRID FEATURES FOR FINGERPRINT RECOGNITION BASED ON INVARIANT MOMENTS AND GLCM

Abstract

The process of verification of the most important processes in security applications and many applications that are special and depends on the identification of the person to ensure privacy. The process of identifying people is done through several things, but the most important of these are biometrics, where people are distinguished by behavioral characteristics (such as gait, signature, sound) or biologic (such as the fingerprint, iris, and vein print).The fingerprint is one of the most important biometrics in humans, because it does not change and be easy to capture in addition to that the human has many fingers, if a problem got a finger can use another finger. In this research a method was proposed to identify the fingerprint by the texture features. Hybrid properties are used by using invariant moments that extract features from a new proposed model is a double thinning model DTM, and GLCM that extracts features from the binary model. The matching process was performed using the Euclidean distance. The system was tested on a database of 40 people. The results were very impressive above 98%. This research opens the horizon to the use of hybrid properties to identify biometrics.