指纹自动识别与人工识别的差异及人机配合策略  

Strategy of Man-machine Coordination by Difference of Computerbased Automatic Fingerprint Identification from Artificial Recognition

作　　者：王勇[1] 吕楠 赵彤 WANG Yong;Lü Nan;ZHAO Tong(Technical Office of Criminal Investigation, Mentougou Branch of Beijing Public Security Bureau, Beijing 102300,China;School of Mathematical Science, University of Chinese Academy of Sciences, Beijing 100049, China;Key Laboratory of Big Data Mining and Knowledge Management, Chinese Academy of Sciences, Beijing 100049, China)

机构地区：[1]北京市公安局门头沟分局刑事科学技术室,北京102300 [2]中国科学院大学数学科学学院,北京100049 [3]中国科学院大数据挖掘与知识管理重点实验室,北京100049

出　　处：《刑事技术》2017年第6期501-503,共3页Forensic Science and Technology

摘　　要：一般而言,技术人员和指纹自动识别系统各自在确定指纹特征、判断两枚指纹相似程度时,采用的方式、方法不完全相同。笔者结合多年的工作经验,对指纹特征提取及指纹特征识别等核心问题展开分析,给出指纹自动识别系统与人工指纹识别在问题判别及处理上的差异之处。在此基础上,针对指纹自动识别系统可能出现特征标注偏差、遗漏特征或者出现虚假特征的情况,总结出人机配合的指纹识别策略。Generally,technicians adopt somewhat different approaches and methods from AFIS(Automatic Fingerprint Identification System)when they independently judge the similarity of two fingerprints based on the determination of fingerprint characteristics.In this paper,the analyses were developed at extracting and recognizing fingerprint features according to our experience,therefore revealing the difference on fingerprint query and processing between AFIS and artificial recognition.Thus,the strategy of man-machine coordination has been concluded against the incidence where AFIS made the fingerprint features labeled with deviation,left out and even erred.

关 键 词：模式识别 指纹识别 人机协同 比对效率 

分 类 号：DF794.1[政治法律—诉讼法学]