Distance based verification techniques for online signature verification system

This paper presents different distance based verification techniques for online signature verification system. For the purpose of extracting the features of the signature, histogram feature extraction technique is used. Each signature is symbolized as a feature vector. In case of verification of the online signature three different distance calculation methods: Manhattan, Chebyshev and Euclidean are applied separately and compared. The experiments were performed on the internationally recognized SVC2004 database. Experimental results obtained using the three different distance calculation based verification techniques confirms that out of these three Euclidean distance calculation based verification technique provide more accurate results. Manhattan and Chebyshev distance calculation method yields equal error rate (EER) 8.5% while Euclidean distance calculation method yields EER 6.5% on skilled forgeries.