Fusion Mechanism for Multimodal Biometric System-Palmprint and Fingerprint

Abstract

The major cause for increase in security breaches, fraud is the degraded quality of stored database. Thus, it is of great importance to make highly secure system in every field like military offices. The project developed for Multimodal Biometric System (MBS) based on image processing. The proposed MBS system is designed for applications like authentication of human being where a fingerprint images and palmprint for each individual used for training stage. The images are captured by using designed hardware; cameras used are Nikon Coolpix S01 10.1MP camera and Xpro night vision camera for palmprint and fingerprint resp. These captured images are preprocessed using Image enhancement techniques and Features are extracted by Gaussian kernel with edge detector, Gabor Filter and Principal Component analysis. Quality measures like length of principal lines, angle between them are also found for these above modalities. Quality base fusion as well as feature level fusion of extracted features are done. Later matching is done by using Canberra distance. This paper proposes unique own generated dataset of used modalities which improves system performance and accuracy up to the 90 to 95%.