320 An On-Line Signature Verification Method Based on Multi-Dimensional Feature Matching

In this paper, a signature verification method based on the multi-dimensional feature matching has been introduced. According to the analysis of the signature static curves and the dynamic writing characteristics, the extraction of the global and local features such as the stroke numbers of signature, the inflection pattern, the time of signature, the writing rhythm, the strength of signature, the changes of intensity and some others has constructed an online signature essential feature representation model. At the same time, based on the probability density function, the threshold obtainment and the signature matching have been also conducted. The experimental results have shown that this method can effectively determine the authenticity of the signature.