基于小训练样本和纹理分析的笔迹鉴别方法

本文针对高等教育自学考试考生试卷笔迹真伪鉴定应用，利用人工笔迹鉴定专家知识，结合文本独立，和训练样本少的特点给出一种基于纹理的算法。通过实验得出，正确接受率为92．9％，正确拒绝率为90．0％。

Writer Identification Based on Small Amount of Test Samples and Texture Analyse

This paper is presented for identification of examination papers handwriting, A kind of algorithm, which is characterized with less training samples and text-independent, is proposed, and it is used of artificial handwriting identification expert knowledge. Finally, experiments show that the correct acceptances rate is 92.9% and the correct rejections rate is 90.0%.