基于智能移动终端的动态签名识别

为在移动终端上进行动态签名识别，采用基于极角特征匹配的动态签名识别方法. 该方法利用签名质心和黄金分割质心建立极坐标系，将从移动终端上获得的直角坐标转换为极坐标后提取极角特征的极值点，并以此为依据分隔签名笔段，通过计算样本与模板之间各个笔段的相似度得到签名总体的相似度，从而判断签名样本与模板是否出自同一人之手. 实验结果表明，该方法的平均误拒率为6.93%，误纳率为11.26%，能够有效地应用于移动终端上的动态签名识别. 

Dynamic Signature Recognition Based on Smart Mobile Devices

In order to recognize the dynamic signature on mobile terminals, a method based on the match of polar angle characteristics was applied. Firstly, a polar coordinate system was established under the centroid of signature and the centroid of the golden section. Secondly, the Cartesian coordinates which were acquired from mobile terminals were transformed to polar coordinates. Then the extreme points of polar angle characteristics were extracted and used to separate the whole signature. Finally, the integral similarity was obtained between the sample signature and template signature by calculating the similarity of all parts of these two signatures, which can be used to judge whether the sample signature and template signature were made by the same person. The experiment results show that the average false rejection rate of the method is 6.93% while the false acceptance rate is 11.26%, so the method can be applied to recognize the dynamic signature on mobile terminals effectively.