指纹汗孔证据评估中的似然比和统计学习方法  

Evaluation of Fingerprint Sweat Pore Evidence Based on Likelihood Ratio Method

作　　者：田润之 赵雅彬[1] 李康[1,2] 崔恒建[3] TIAN Run-zhi;ZHAO Ya-bin;LI Kang;CUI Heng-jian(School of Investigation,People's Public Security University of China,Beijing 100038,China;Zhejiang Police College,Hangzhou 310053,China;School of Mathematical Science,Capital Normal University,Beijing 100048,China)

机构地区：[1]中国人民公安大学侦查学院,北京100038 [2]浙江警察学院,浙江杭州310053 [3]首都师范大学数学科学学院,北京100048

出　　处：《数理统计与管理》2025年第1期1-19,共19页Journal of Applied Statistics and Management

基　　金：国家自然科学基金资助项目(12031016,11971324);公安部技术研究计划(2019JSYJC20)。

摘　　要：指纹在人身认定中具有非常重要的作用,而对于指纹三级特征如何应用于指纹鉴定中是制约指纹技术发展的一个难题。针对这一问题,本文以1200dpi分辨率下的指纹图像为样本,利用欧氏距离相似度计算汗孔特征组之间的相似性,结合评分法似然比的计算以及评价模式,比较了边缘似然比、联合似然比、支持向量机以及T-sne方法,最后比较了不同汗孔数量下,似然比模型的精度变化情况,结果表明联合似然比可以获得相对较低的错误率,同时随着所标注汗孔数量的减少,似然比模型的错误率也在不断提高。本文将指纹汗孔特征与似然比相结合,深入探究了应用似然比方法评估指纹汗孔特征的证据价值的可能性,以推进指纹三级特征应用于指纹鉴定工作中的发展进程。Fingerprints play a very important role in the identification of persons,and the application of fingerprint third level characteristics to fingerprint identification is a difficult problem that limits the development of fingerprint technology.To solve this problem,this paper uses Euclidean distance similarity to calculate the,similarity between sweat pore feature groups,and compares marginal likelihood ratio,joint likelihood ratio,support vector machine and T-sne method by combining the calculation and evaluation of likelihood ratio method,with a fingerprint image at a resolution of 1200dpi as a sample.Finally,the accuracy changes of likelihood ratio model under different number of sweat pores are compared.The results show that the joint likelihood ratio was better than the other method,and as the number of marked sweat pores decreased,the false rate has got higher.In this paper,we combine fingerprint sweat pore features with likelihood ratio,and explore the possibility of using likelihood ratio method to evaluate the evidence value of fingerprint sweat pore features.

关 键 词：指纹三级特征 法庭科学 法庭证据评估 欧氏距离相似度 似然比 

分 类 号：DF7941[政治法律—诉讼法学] O212[政治法律—法学]