基于光学体全息和小波包分解的虹膜识别实现 | |
黄欢 ; 才德 ; 严瑛白 ; 金国藩 ; 何庆声 ; HUANG Huan ; CAI De ; YAN Yin-bai ; JIN Guo-fan ; HE Qing-sheng | |
2010-06-08 ; 2010-06-08 | |
关键词 | 虹膜识别 体全息相关 小波包分解 特征图像 Iris recognition Volume holographic correlation Wavelet packet decomposition Eigen image TP391.41 |
其他题名 | Implementation of iris recognition based on volume holographic correlation and wavelet packet decomposition |
中文摘要 | 在光电混合的体全息系统上实现虹膜识别。利用小波包分解生成特征图像,并且穷举找到具有较高识别率的小波包节点组合。耗时的特征图像生成和图像存储都是事先进行的,光学体全息凭借其多通道和高并行性可以实时地完成特征提取和相关。计算机对采集到的相关结果进行后处理,通过选择合适的窗口和归一化可以进一步提高识别率。模拟识别率可达98%,实验中的最高识别率为91%。实验结果证明了方案的可行性,并为今后向实用化方向发展奠定了基础。; An implementation of iris recognition on a hybrid opto-electronic volume holographic system is introduced in this paper. Wavelet Packet (WP) decomposition is used to generate the eigen images. The WP nodes combination with the highest recognition rate is obtained through exhuastive search. The time-consuming stages of eigen image generation and storage are completed preliminarily, and due to its multi-channel and high parallelism ability, the feature extraction is performed in real time with the help of volume holographic correlation. The correlation results detected by CCD are post-processed in a computer. Recognition rates can be further improved if window of proper size and the normalization technique are used. The recognition rate of computer simulation can reach 98%, while the highest rate in experiment is 91%. Experiment results demonstrate the feasibility of the iris recognition implementation scheme, and provide some foundation for future development.; 国家自然科学基金(60277012)-“光学子波包变换模式识别技术” |
语种 | 中文 ; 中文 |
内容类型 | 期刊论文 |
源URL | [http://hdl.handle.net/123456789/48655] ![]() |
专题 | 清华大学 |
推荐引用方式 GB/T 7714 | 黄欢,才德,严瑛白,等. 基于光学体全息和小波包分解的虹膜识别实现[J],2010, 2010. |
APA | 黄欢.,才德.,严瑛白.,金国藩.,何庆声.,...&HE Qing-sheng.(2010).基于光学体全息和小波包分解的虹膜识别实现.. |
MLA | 黄欢,et al."基于光学体全息和小波包分解的虹膜识别实现".(2010). |
个性服务 |
查看访问统计 |
相关权益政策 |
暂无数据 |
收藏/分享 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论