Trace copy forgery detection for handwritten signature verification

Abstract

Handwritten signature verification has been extensively studied in past decades. Its many applications include banking, credit card validation, security systems etc. We mainly deal with the trace copy forgery detection problem in a handwritten signature verification system. Based on one of our previously study [P. S. Deng et al. (1999)], we have proposed an efficient approach, called Deng's approach to build-up an offline handwritten signature verification system by using wavelet transformation technique. From the experimental results, the accuracy rate of the Deng's approach [P. S. Deng et al. (1999)] is 92.57% and 93.68% for English and Chinese respectively. The approach we proposed not only got a good accuracy result, i.e. 87%, but also has many other advantages as follows: (1) it can be applied in many languages including English and Chinese, (2) it can be adopted to both on-line and off-line applications, and (3) it uses all the strokes information rather than only the vertical strokes information in a Chinese signature.