Zernike Moments-based Offline Handwritten Signature Verification Research

The signature is a traditional way of identity recognition, which has a legal status that can't be replaced. On a lot of occasions, signature represents the signer himself. With the fast development of information technology, the method of the traditional direct recognition can't adapt to present application's needs gradually. How to recognize the truth or falseness of the handwritten signature using computer automatically becomes the urgent subject to be solved. Therefore signature verification has good application prospects and enormous commercial value. Lots of scholars and research institutions all over the world have already demonstrate the great interest and do a large number of jobs. But so far, the research results of already obtained also have a very great distance to this question's settlement completely. With the rapid development of national economy, , the importance of signature verification is obvious day by day in the office automation and public safe field. Therefore it is very essential to carry out the research work of signature verification.The main difficulty in offline handwritten Chinese signature verification is feature extraction. Validity of feature decides the performance of signature verification system directly. The feature extraction is a key point which this article studies. On the basis of analyzing and summarizing the method of off-line handwritten signature verification, this article have proposed a method of feature extraction based on that Zernike moment and Gabor transformation. We proved validity and superiority of this new feature extraction method through the experiment. Zernike moment can descript the global feature of pattern well, it has rotation invariance. And the Gabor transformation is one kind of Short Time Fourier Transform in fact, which have excellent performance in analyzing the frequency and direction information locally in the digital image. So we extract global feature of signature with the orthogonal Zernike moment, extract local feature of signature with Gabor transformation. The feature which is extracted...