相似度算法在手写签名质谱成像鉴定中的应用研究  

Studies on verification of handwritten signatures by mass spectrometry imaging with the Similarity Algorithm

作　　者：刘亚丽[1,2] 贾滨[2] 丁丽英[2] 吴佳[2] 陈焕文[2] 章文军[1] 

机构地区：[1]河北工业大学化工学院,天津300130 [2]东华理工大学/江西省质谱科学与仪器重点实验室,江西南昌330013

出　　处：《计算机与应用化学》2012年第5期541-544,共4页Computers and Applied Chemistry

基　　金：国家重大科学仪器设备开发专项(2011YQ170067);国家自然科学基金资助项目(21175019)

摘　　要：质谱成像技术无需任何样品预处理,即可获取待测样品的分子信息和分布情况。本文采用表面解吸常压化学电离质谱(SDAPCI-MS)技术对手写签名样品进行检测,通过对所得的质谱特征峰信号进行成像处理,获取书写油墨分布的强度信息。实验结果表明真实签名和伪造签名因为笔压轻重不同而油墨分布位置不同,据此能够区分签名的真伪。应用相似度算法对手写签名的特征成像数据进行分析,在样品量较少的情况下,能够客观的比较真迹之间以及真迹和伪迹之间的相似程度,进而实现对真伪签名的准确鉴定。本方法操作简便,耗时较短,对手写笔迹的鉴定具有重要的借鉴意义。Without any sample pretreatment, mass spectrometry imaging （MSI） can provide molecular information and the molecular distribution information on the sample. Handwritten signature samples were directly detected by surface desorption atmospheric pressure chemical ionization mass spectrometry （SDAPCI-MS）. The mass spectra recorded by SDAPCI-MS were subjected to data processing. It is easy to distinguish the authentic signature and the forged signature by observing the strength of the corresponding mass spectrometry imaging of the handwritten signature. Considering the relative ink concentration feature, which is extracted from the data of the handwritten signature, as a feature of the signature, we provide useful data by processing the data gained from mass spectrometry imaging. The Similarity Algorithm is applied to the handwritten signature verification, and this method gives excellent results. The results demonstrate that this method is both simple and fast, and the method is a promising technique for the verification of handwritten signatures with few samples.

关 键 词：质谱成像 相似度算法 手写签名 

分 类 号：TQ015.9[化学工程] TP391.9[自动化与计算机技术—计算机应用技术]