计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (30): 162-166.

• 数据库与信息处理 • 上一篇    下一篇

基于隶属度函数的决策层融合算法

叶学义1,赵知劲1,张文聪2,赵泽茂1   

  1. 1.杭州电子科技大学 通信工程学院 模式识别与信息安全实验室,杭州 310018
    2.中国科学技术大学 电子科学与技术系 智能信息处理实验室,合肥 230027
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-10-21 发布日期:2007-10-21
  • 通讯作者: 叶学义

Decision-layer fusion method based on subjected function

YE Xue-yi1,ZHAO Zhi-jin1,ZHANG Wen-cong2,ZHAO Ze-mao1   

  1. 1.Lab of Pattern Recongnition and Information Security,Hangzhou Dianzi University,Hangzhou 310018,China
    2.Lab of Intelligent Processing,University of Science and Technology of China,Hefei 230027,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-10-21 Published:2007-10-21
  • Contact: YE Xue-yi

摘要: 虹膜识别对所采集的样本图像较为苛刻的要求所造成的高采集失败率和注册失败率是阻碍它推向大规模应用的重要原因,也是目前虹膜识别研究的难点问题之一。在适当降低虹膜图像采集质量要求的条件下,能够大幅降低采集失败率和注册失败率,但同时也降低了识别的性能。采用数据融合的方法,在识别的决策层引入对应的脸像特征模板的相似度判据,根据所提出的基于隶属度函数的决策层融合算法实现融合决策。实验结果表明,这种多生物特征融合决策的方法既可以使得虹膜的采集失败率和注册失败率保持在实际应用可以接受的程度,又能够使得整体的识别性能达到目前虹膜识别技术的水平。

Abstract: The high rate of failure to capture or to enrollment of iris recognition is one of the most important reasons that hamper people to broaden this technology to a large scale real application,because of the rigorous requirement to iris images used for recognition,and the big challenge to researchers.If properly releases some limits capturing iris images,the disadvantage will be improved famously,but meantime some good performances of the iris recognition will be depressed.From the perspective of data fusion,here proposes the subjected function based decision-layer fusion approach to realize the fusion identification by joining the parameter of the facial similarity.Experimental results show that this multi-biometrics fusion identification approach not only has the failure rate to capture or to enrollment acceptable to the real application of the iris recognition,but also preserves present eminent performances of the iris recognition.