On-line chinese signature verification algorithm based on dynamic time warping and hidden Markov model

Signature verification is one of the most important research areas in the field of personal identification through biometrics technology. The technology is also regarded as a front subject in many fields such as signal,processing, information sensoring, pattern recognition and information security. On-line signature verification schemes extract signature features which characterize spatial and temporal characteristics of a signature. In this paper, we present a novel on-line handwritten signature verification algorithm using DTW (Dynamic Time Warping) and HMM (Hidden Markov Model), and offer the definition and algorithm of the signature model. Furthermore, the paper puts forward the self-learning and recognition methods of the model. The optimized algorithm maintains the viewpoint of the state of HMM and overcomes the disadvantages of DTW, therefore it's more suitable for the universal situations. Our database includes 867 genuine signature and 1754 forgery signatures which were collected from 126 volunteers. The comparative experiment results show that in signer dependent situation, the performance of the united model is equivalent to DTW and better than HMM, and in signer independent situation, it is better than both DTW and HMM.