学位论文-基于手部特征的多模态生物识别技术研究.docVIP

基于手部特征的多模态生物识别技术研究 基于手部特征的生物多模态技术研究 摘要 关键词　 Based on the biological characteristic of hand multi-modal technology research Abstract With the pace of science and technology advancing with the times, people are more and more concerned about security issues. Biological recognition technology advantages more prominent its importance, biological feature identification is to identify the physiological characteristics and behavior characteristics of human beings when they are used for identification. But at the same time, there are some disadvantages in the single mode based identification technology, for example, poor security performance, easy to be stolen, etc. In this paper, we put forward a multi modal biometric technology. It is through the use of a variety of biological features of the fusion technology for identification and authentication, major advantages in improving the recognition efficiency, noise immunity, universality and other aspects have been improved. Although in recent years the modal biometric technology has been more and more attention of all levels. However, the technology is still not very mature, we need further research and discussion in future work. We will have rich characteristics of the human hand as the focus of this paper, The hand biometric is mainly based on Palmprint and hand geometry, they are as a biometric characteristic with stability and uniqueness, this paper is mainly based on wavelet transform analysis on biological characteristics of proposed biometric ,The extraction method of biological features is based on Wavelet Transform, then puts forward the improved method of canonical correlation analysis, Two kinds of fusion method of feature vector and palmprint features in the hand shape biological feature level, this method is mainly by introducing the correction coefficient, to adjust the projection direction of the feature vector, the criterion function value reached the maximum. In this paper, The use of b