Handwritten Signature Veri.cation (HSV) is a discipline which aims to validate the identity of writers according to the handwriting styles. Compared with on-line HSV, off-line HSV is less limited in equipment involvement and can be applied in more .elds. Nevertheless, it is more dif.cult to manipulate than on-line HSV due to the loss of dynamic information during the writing process, such as writing position, velocity, acceleration and pressure. In this paper, we focus on off-line HSV and present a new feature selection method based on Contourlet, which gives full play to the merits of both conventional structure feature and statistical feature. After dimensionality reduction to extracted eigenvector by K-L transform, genuine signatures and forgeries are distinguished through support vector machines (SVM). The result of our experiment has con.rmed the effectiveness of the proposed approach.
HSV Contourlet SVM K-L Transform
Ming Yang Zhongke Yin Zhi Zhong Shengshu Wang Pei Chen Yangsheng Xu
School of Information Science and Technology, Southwest Jiaotong University, Chengdu, China;Shenzhen School of Information Science and Technology, Southwest Jiaotong University, Chengdu, China Department of Automation, Shanghai Jiao Tong University, Shanghai, China;Department of Mechanical an Shenzhen Institute of Advanced Integration Technology, CAS/CUHK, Shenzhen, China;Department of Autom Shenzhen Institute of Advanced Integration Technology, CAS/CUHK, Shenzhen, China Shenzhen Institute of Advanced Integration Technology, CAS/CUHK, Shenzhen, China;Department of Mecha