Unimodal biometric systems have involved a range of researchers and achieved great success. Unimodal system alone may not be able to meet the increasing requirement of high accuracy i n today?s biometric system. Unimodal biometric systems suffer from much challenge such as noisy data, non-universality and spoof attacks. Multimodal biometric systems can resolve these limitations effectively by using two or more individual modalities. In this technique fusion of iris, fingerprint and face traits are used in order to improve the accuracy, security of the system and to identify the human. The main purpose is to come across over whether the combination of fingerprint, iris and face biometric can achieve performance that may not be possible using a single biometric technology. Local ternary pattern Extraction and Feature Selection Using GLCM algorithm with SVM classifier are used for finger, iris and face images. This system offer the high performance and to overcome the limitation of single modal biometrics. This new proposed system is producing more reliable results than the 

existing method.