Tri modal biometrics systems are becoming increasingly efficient over the unimodal system, especially for the securing mobile devices like PCA, PC, tablets, etc. In this project, we propose a bi-modal biometric recognition technique using finger and voice as biometric traits. This project addresses the issues of score fusion techniques in Tri imodal biometrics verification systems. Fusion techniques namely weighted sum rule, KNN classifier and support vector machine (SVM) has been evaluated with the matching scores of the two biometrics modalities namely finger and voice. The experiments with the fusion techniques were conducted over a BioChave database collected from 10 individuals with multiple instances of the two traits. Experimental results showed that SVM rule gave the best performance among the fusion techniques. Hence, we confirmed that the proposed Support Vector Machine fusion method outperformed other fusion techniques and unimodal classifiers.