摘要
本文针对在线汉字签名鉴别系统中存在的误拒绝真实签名问题和误接收伪造签名问题,提出了相应的解决方法。首先引入伪匹配起点的概念,使得用动态规划算法求出的最优匹配路径更具合理性;采用了多模板的方法,认为最终的鉴别距离是被测签名与多个参考签名距离的平均值;根据最小风险贝叶斯决策规则,求出距离阈值h使系统的平均风险达到最小。实验结果证明了本文提出方法的有效性。
To deal with the problem of false rejection of genuine signatures and that of false acception of forgery signatures, some solutions are proposed in this paper. First, the pseudo matching starting point is introduced in order to make the optimal matching path solved by dynamic programming algorithm more reasonable. Then, multi-templates are used and the final verification distance is defined as the average distance between the detected signature and several reference signatures. Last, based on the Bayesian minimum risk decision rule, the distance threshold h which makes the average risk minimun is found. The effectiveness of the proposed solutions is demonstrated by the experiments.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2000年第4期451-455,共5页
Pattern Recognition and Artificial Intelligence
关键词
动态规划法
汉字签名鉴别
序列匹配
模式识别
On-line Chinese Signature Verification, Dynamic Programming Algorithm, Bayesian Minimum Risk Decision