To identify or authenticate a person using bimodal biometric system.The biometric features are given as input. Voice is recorded through microphone and images for palm are obtained using palm vein & vein scanner.
- To overcome the drawbacks of the unimodal system with more than one modality\
- To eliminate the False Rejection Rate(FRR) and False Acceptance Rate(FAR)\
- To authenticate a person accurately using the provided input biometrics\
- To restrict the access to a system by unauthorized individuals\
- Speech - MFCC feature are extracted for each samples and stored in a pickle format. The MFCC features are loaded from the pickle file to a numpy array.
- Palm Image - The palm image features are extracted by 2D Gabor feature extraction method using opencv library
The extracted features are trained using Support vector machine.