Finger Knuckle Print Recognition Using Complex Conjugate Feature Vector

Abstract

AbstractBiometrics are human-specific traits that are employed in person identification and access control. Various biometrics such as fingerprints, finger knuckle, iris, palmprint, vein patterns, and DNA are used in recognition. Among these numerous biometrics, the researcher is drawn to the hand-based finger knuckle print (FKP). On the dorsal side of the hand, the FKP biometric is found. The creases and folds in the finger knuckle print are rich in textural pattern and can be utilized to identify individuals. The recognition of finger knuckles based on a complex feature fusion is proposed in this chapter. The subspace techniques such as principle component analysis (PCA) and linear discriminant analysis (LDA) are used to extract complex number features. These extracted complex vectors are fused using a parallel fusion strategy. Finally, finger knuckle PolyU and IIT Delhi datasets are used to test the developed parallel fusion complex features. The experimental results show that the proposed parallel fusion of complex vector for feature extraction techniques improve the classification accuracy.KeywordsFeature extractionPCALDARecognitionFinger knuckle print