生物特征信息处理开发包BITK及其应用平台  

Biometric Information Processing Toolkit and Its Application Platform

作　　者：何晓光[1] 田捷[1] 何余良[1] 杨鑫[1] 

机构地区：[1]中国科学院自动化研究所,北京100080

出　　处：《计算机仿真》2007年第8期185-190,共6页Computer Simulation

摘　　要：由于人类固有的生物特征能用来进行可靠的身份认证和识别,近10年来生物特征识别已经成为研究热点。为了对生物特征识别研究提供软件支持,设计研发了生物特征信息处理开发包(Biometric Information Processing Toolkit,BITK)。BITK是一个C++软件开发包。它以非线性数据流模型来组织整体计算流程,提供了一个可扩展、可重用的框架来整合生物特征识别领域的算法和数据结构。BITK还设计了一个精干的可视化框架以及一个管理各种生物特征采集仪的一致性框架。此外,在BITK基础上研发了生物特征信息处理平台(BITK-based Application Platform,BITKAPP)。该平台通过灵活的plugin架构和友好的用户界面,充分发挥BITK的功能,并降低了BITK的使用门槛。在实际应用中,BITK和BITKAPP向研究人员提供了一套高效率的研究、实验及开发手段,并且作为支撑平台完成了第一届生物特征识别竞赛(the 1st Biometric Verification Competition)的多模态数据采集任务。Biometrics is a reliable technology for personal authentication and identification, more and more researches are focused on it in the recent decade. To effectively support Biometrics research in terms of software, Biometric Information Processing Toolkit （BITK） is designed and implemented. BITK is a C ＋＋ software development toolkit （SDK）. Its computational framework is designed with the non - linear data flow model. And it provides an extensible and reusable framework to integrate algorithms and data structures of Biometrics. Moreover it designed a refined visualization framework and a consistent biometric sensor managing framework. In addition, an application platform based on BITK （BITKAPP） is developed. With the friendly user interface and the dynamic plug - in architecture, BITKAPP makes the best of BITK while easing the manipulation of BITK. In practices, the toolkit and platform effectively support the biometrics research, experiment and implementation. And as supporting software, they have constructed the multi - mode database for the 1 st Biometric Verification Competition.

关 键 词：非线性数据流模型 可扩展框架 一致性框架 生物特征识别开发包 生物特征识别平台 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]