查看文章

In this paper we focus on constructing k-Nearest Neighborhood Structure(k - NNS) for minutiae points in a fingerprint image. For each minutiae point in a fingerprint, a k - NNS is constructed taking the local and global features of minutiae points. This structure is quantized and mapped onto a 2D array to generate a fixed length 1D bit-string. Further this bit string is applied with a DFT to generate a complex vector. Finally the complex vector is multiplied by a user specific random matrix to generate the cancelable template. We tested our proposed method on database FV C2002 and experimental results depicts the validity of the proposed method in terms of requirements of cancelable biometrics namely diversity, accuracy, irreversibility and revocability.