多技术合力的人脸识别系统设计  

Face recognition system designed to integrate multiple techniques

作　　者：苏光大[1] 

机构地区：[1]清华大学电子工程系,北京100084

出　　处：《智能系统学报》2009年第6期471-474,共4页CAAI Transactions on Intelligent Systems

基　　金：国家"十一五"科技支撑计划资助项目(2006BAK08B07)

摘　　要：人脸识别系统结构复杂,也包含了多种技术.从多个角度阐述了多技术合力的人脸识别系统设计方法,内容涉及人脸图像采集、人脸识别算法、并行处理、综合系统集成4个部分.多技术合力的人脸识别系统设计方法体现了速度与智能化相结合的优势,也体现了系统的完备性和性能的互补性.给出了该方法在人脸识别率和识别速度上达到的性能指标,并给出了所设计的人脸识别系统的部分突出应用.Face recognition systems usually have complicated structures and integrate many techniques.In this paper,the design of a face recognition system that combines various technologies was explained from different perspectives,including facial image capture,face recognition algorithms,parallel processing,and system integration.Multi-technique integration reflects the advantages of combining hardware speed with artificial intelligence,as well as the completeness and complementary performance of the proposed face recognition system.Experiments with the suggested system showed improvements in recognition rate and recognition speed.Finally,some special applications of this system were presented.

关 键 词：人脸识别 GABOR滤波 图像并行处理 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]