ABSTRACT

Several biometric modalities are currently being tested for identity verification, but amongst all the possible biometric modalities, the handwritten signature has been used for the longest period of time as a means of identification. It is commonly found in commerce and banking transactions, credit card payments, cheque authentication and, in general, all types of legal documents. Therefore, considering all the different biometric modalities, the signature is undoubtedly the most accepted for the majority of different scenarios. Advancing progress in identification applications has led to widespread demand for new generation ID documents, such as electronic passports and citizen cards, which contain additional biometric information required for more accurate user recognition. The image of the users handwritten signature is already incorporated into ID documents. However, current error rates in verifying signature images are not yet sufficient for massive deployment .This can be overcome by embedding dynamic features of signature along with the static features within the documentation. This problem and the increasing demand for standardized signature verifications systems have motivated the research work performed in present study. Accuracy of the system depends on how these dynamic features are extracted. In literature several methodologies have been given to extract these features and since this field of signature verification is still under development phase, many methodologies are yet to be explored. One such unexplored methodology based on Discrete Fractional Cosine Transform is presented in this study. There is a close relationship between the conventional Discrete Cosine Transform (DCT) and the Discrete Fractional Cosine Transform (DFrCT).The DFrCT share many useful properties of the regular cosine transform, and has a free parameter, its fraction. When the fraction is zero, we get the cosine modulated version of the input signal. When it is unity, we get the conventional DCT. As the fraction changes from 0 to 1 we get different forms of the signal which interpolate between the cosine modulated form of the signal and its DCT representation. Thus, DFrCT is a general form of DCT which has an additional free parameter, and with this free parameter it may find its place in many applications more efficiently as compare to where DCT is found to be useful. A new method for an on-line handwritten signature verification based on finite impulse response (FIR) system is proposed by utilizing discrete fractional cosine transformation (DFrCT) for

feature extraction. Various characteristics of the hand-written signature are used to extract different features of the signature by optimizing the value of the fractional order. The system for the hand-written signature verification is realized by characterizing three FIR systems. The impulse responses of FIR systems are used to calculate Euclidean norm. The signature can be verified by evaluating the difference between the average of Euclidean norms of reference signatures and the Euclidean norm of signature to be verified. The equal error rate (EER) is calculated to compare the efficiency of the proposed method. It has been verified through simulation results that the DFrCT tool achieves much better results as compared to discrete cosine transform (DCT) for extracting the features. The signature verification experiment was performed on SVC2004 signature database.