Writer-independent Offline Signature Recognition based upon Fourier Descriptors

Signatures of individuals play an important role in financial, commercial and legal transactions and hence require a secured recognition and verification system. This paper presents a novel approach towards signature recognition using Fourier Descriptors. The signature image is enclosed in a closed contour and the Fourier terms of these contours are computed by representing the contour in a complex plane. The FDs computed are then fed to K-NN classifier for recognition. The proposed system exhibits good recognition accuracy.