CAJ | 학술논문

笔迹鉴别是通过分析手写笔迹信息来判断书写者身份的特征分析技术，为提高笔迹鉴别的科学性和准确性，设计了基于ARM嵌入式系统的文本独立（Text-independent）笔迹鉴别系统。系统包括笔迹采集端、数据传输部分和上位机处理端3个部分。首先给出了总体设计和详细的工作流程，并对笔迹信息采集模块和数据传输部分进行分析，最后结合笔迹图像预处理技术和纹理特征分析方法，采用基于多维度Gabor变换和支持向量机（Support Vector Ma-chines，SVM）方法进行实时笔迹鉴别。实验表明，系统鉴别率高，达到了良好的预期效果。
필적감별시통과분석수사필적신식래판단서사자신빈적특정분석기술，위제고필적감별적과학성화준학성，설계료기우ARM감입식계통적문본독립（Text-independent）필적감별계통。계통포괄필적채집단、수거전수부분화상위궤처리단3개부분。수선급출료총체설계화상세적공작류정，병대필적신식채집모괴화수거전수부분진행분석，최후결합필적도상예처리기술화문리특정분석방법，채용기우다유도Gabor변환화지지향량궤（Support Vector Ma-chines，SVM）방법진행실시필적감별。실험표명，계통감별솔고，체도료량호적예기효과。
Writer Identification is one of the characteristics technology to determine the writer's identity by analyzing handwriting information. In order to improve the accuracy and scientificity of writer identification, a text-independent writer identification system based on ARM embedded system is developed. This system includes three parts: handwriting capture terminal, data transfer terminal and PC processing terminal. Firstly, the overall design and detailed workflow of the system is described. Meanwhile, the handwriting capture terminal and data transfer terminal is analyzed. Finally, combined with handwriting image preprocessing techniques and texture analysis methods, a real-time writer identification is designed by multidimensional Gabor transform and Support Vector Machines methods. According to the experiments, the system has high identification rates and achieve the desired good effects.