摘要: 针对掌纹识别在智能移动设备上的应用,提出一种掌纹识别算法,并嵌入到联想ET980智能手机,开发基于移动设备的掌纹验证系统。以新的方式采集掌纹图像并给出相应的预处理算法。使用优化的Gabor判别方法提取掌纹特征,减少特征提取的时间。系统验证一张掌纹图像的时间小于1 s,同时可获得EER=3.89%的识别精度,满足验证系统的实时性和精度要求。
关键词:
掌纹验证系统,
移动设备,
掌纹分割,
光照规一化,
优化Gabor-LDA特征
Abstract: Aiming at the application of palmprint recognition in mobile devices, this paper proposes a palmprint recognition algorithm and successfully embeds it into Lenovo ET980 mobile phone. The mobile-based system captures and pre-processes palmprint image in a new style. An optimal Gabor wavelets based discriminant subspace method is applied for feature extraction to give an effective and compact presentation of palmprint image. The system achieves the recognition rate of EER=3.89% and takes 1 s to verify a palmprint image, which satisfies the requirement of a real-time verification system.
Key words:
palmprint verification system,
mobile device,
palmprint segmentation,
illumination normalization,
optimal Gabor-LDA feature
中图分类号:
张建新;欧宗瑛;刘典婷. 基于移动设备的掌纹验证系统[J]. 计算机工程, 2010, 36(4): 164-165.
ZHANG Jian-xin; OU Zong-ying; LIU Dian-ting. Palmprint Verification System Based on Mobile Device[J]. Computer Engineering, 2010, 36(4): 164-165.