一种静态手写汉字签名特征描述算法的实现  

The Realization of the Described Algorithm of a Static Handwritten Chinese Character

作　　者：王洪革[1] 宋晓雪[2] 潘石[2] 

机构地区：[1]吉林警察学院信息工程系,吉林长春130117 [2]吉林大学通信工程学院,吉林长春130012

出　　处：《软件导刊》2012年第9期43-45,共3页Software Guide

摘　　要：基于模式识别统一信息熵理论,采用一种静态手写汉字签名特征描述算法,选取静态手写汉字签名样本的两个特征,利用DB8小波对签名样本进行小波分解,对主要能量分布方向进行由粗到细的三级筛选,得到当事人和待鉴定人签名样本整体和某一特征的差异度,并且,当事人主要特征上的差异度大小得以保留。该特征描述方法可应用于静态手写汉字签名的区别鉴定。Based on the pattern recognition unified information entropy theory, using a static handwritten Chinese charac- ter signature description algorithm, the static handwritten Chinese character signature sample two characteristics, the use of DB8 wavelet for signing sample wavelet decomposition, the main energy distribution direction from coarse to fine tertia- ry screening, get parties and stay appraiser signature of the whole sample and a certain character difference degree. From the results, the parties concerned main characteristic differences degrees size preserved, this description method can be applied to static handwritten Chinese character signature difference identification.

关 键 词：信息熵 特征提取 静态手写汉字签名 特征描述 小波分析 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]