实验室资源安全管理策略——序列化多生物特征识别系统框架  

Strategy for safety management of laboratory resources:Framework of serialized multi-biometric characteristic identification system

作　　者：张擎[1] 刘淑美[1] Zhang Qing;Liu Shumei(School of Fine Art,Shandong University,Ji’nan 250100,China)

机构地区：[1]山东大学艺术学院,山东济南250100

出　　处：《实验技术与管理》2018年第11期249-253,267,共6页Experimental Technology and Management

基　　金：山东省高校证据鉴识重点实验室项目"基于生物特征识别的痕迹检验技术研究"(KFKT(SUPL)-201709)

摘　　要：在实验资源的安全管理工作中防止非法资源获取是关键,而合法身份认证是核心技术。不同于传统密码、IC卡等身份认证技术,生物特征识别具有不会丢失、遗忘、不易被篡改的优势。该文分析实验室资源安全管理的特殊要求,提出一种多生物特征识别系统框架,结合并行融合和序列化融合方案,利用在线的半监督学习技术提高生物特征的识别性能。经实验证实,该框架在安全性和使用方便性方面具有优势,更加符合实验室资源安全使用和管理的需求。Prevention of illegal resource acquisition is the key to the safety management of experimental resources,and legal identity authentication is the core technology.Unlike traditional authentication technologies such as the password,IC card,etc.,the biometric recognition has the advantages that it is not lost,forgotten and not easily tampered with.This paper analyzes the special requirements for the safety management of laboratory resources and proposes a framework for the multi-biometric characteristic identification system.In combination with the scheme of the parallel integration and serialization integration,the identification performance of biological features is improved by using the on-line semi-supervised learning.Experiments show that this framework has advantages in safety and easy utilization and is more suitable for the safe use and management of laboratory resources.

关 键 词：实验室资源安全管理 多生物特征识别 序列化集成 半监督学习 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]