Multimodal Biometric Recognition System Using Feature Level Fusion

With so many recent advancements in security, there is no doubt that biometric recognition will take over everything. This paper is to list a few ways of combining multiple biometric modals using fusion. The paper mainly focuses on multiple processes used for feature extraction from face and palmprint. FACE 94, 95 and 96[1] are face databases and IITD palmprint database is used for palm. The features were extracted using various methods like Gabor, Radon, Ridgelet and Radon-Gabor. Feature level fusion was applied using algorithms like Mean, weighted average. For the classification SVM classifier was used. Another method used for classification was confidence based similarity checking using SSIM.