基于纹理分析的笔迹鉴别系统

Writer Identification System on Texture Analysis

笔迹鉴别是通过分析手写字符的书写风格来判断书写人身份的一门技术。本文把手写笔迹作为一种纹理来看待,将笔迹鉴别转化为纹理识别来处理,利用多通道Gabor滤波器来提取笔迹图像的纹理特征,用支持向量机进行分类。实验中采集了17个人的不同笔迹,取得了较好的结果。

Writer identification is a kind of technology by analyzing written style to identify the authors. This text takes the handwriting as a kind of texture, transform the writer identification for texture identification, withdraw the texture characteristics of handwriting images by multi-channel Gabor filtering, carry on the classification with the support vector machines. In the experiment, we gathered different handwritings from 17 people and obtained a better result.