LI Jin-ming1, MA Lin1, ZHANG Shao-hua2, CHENG Ya-li1, CHENG Nai-peng1
(1. School of Instrument and Electronics, North University of China, Taiyuan 030051, China; 2. Shanghai Aerospace Electronic Technology Institute, Shanghai 201109, China)
Abstract： Due to complex computation and poor real-time performance of the traditional iris recognition system, iris feature is extracted by using amplitude and phase information of the mean image blocks based on Gabor filtering on image, and the k-nearest neighbor algorithm is combined to complete iris recognition function. The recognition reduces the recognition time and improves the recognition accuracy. At the same time, identification result is transmitted to the cloud server through ZigBee network to solve diffcult wiring problem. The experiment shows the system runs stably and has fast recognition speed. It has been applied to a security system.
Key words： iris recognition; digital signal processor (DSP); ZigBee; image block
CLD number： TP391.41 Document code： A
Article ID： 1674-8042（2018）02-0169-05 doi： 10.3969/j.issn.1674-8042.2018.02.011
References
［1］Chen Y R. Research on intelligent computing and biometric identification technology. Journal of Tianjin University of Science & Technology, 2015, 30(4): 78.
［2］Li H L, Guo L H, Wang X Z, et al. Iris recognition based on weighted Gabor filter. Journal of Jilin University (Engineering and Technology Edition), 2014, 44(1): 196-202.
［3］He Z H. Iris recognition based on generalized Gaussian distribution FDCT_Wrap and FSVM. Telecommunications Science, 2016, 32(7): 126-131.
［4］Fang Q, Yao P. Reliable iris recognition using 2D quadrature filters. Computer Science, 2015, 42(5): 281-285.
［5］He M S, Luo W S, Chen Q X et al. Research on anti-fatigue driving alarm system by iris recognition based on DSP. Journal of Safety Science and Technology, 2016, 12(1): 127-131.
［6］He F. Research on the iris recognition algorithms of multiple feature extraction and fusion based on Gabor filtering. Jilin: Jilin University, 2015.
［7］Zhang Z, Shao X X. A SIFT iris matching algorithm. Journal of Zhengzhou University (Natural Science Edition), 2017, 49(3): 14-19.
［8］Hu P, Wu Y B, Cao D H. Interest point extraction algorithm for robotic grasp and implementation on DSP. Instrument Technique and Sensor, 2015, (11): 85-88.
基于DSP和ZigBee的虹膜识别系统设计
李锦明1, 马林1, 张少华2, 成雅丽1, 成乃朋1
（1. 中北大学 仪器与电子学院， 山西 太原 030051； 2. 上海航天电子技术研究所， 上海 201109）
摘要:针对传统虹膜识别系统计算复杂、 实时性差等问题， 在Gabor滤波图像的基础上， 利用分块图像的幅值均值和相位信息来提取虹膜特征量， 并结合k-近邻算法来完成虹膜识别功能， 缩短识别时间， 提高识别准确度。 同时， 将识别结果通过ZigBee网络发送至云端服务器， 解决了厂区内布线困难问题。 实验表明， 该识别系统运行稳定， 识别速度快。 该识别系统已被应用于某安防系统。
关键词:虹膜识别； 数字信号处理器； 紫蜂协议； 图像块
引用格式：LI Jin-ming, MA Lin, ZHANG Shao-hua, et al. Design of iris recognition system based on DSP and ZigBee. Journal of Measurement Science and Instrumentation, 2018, 9(2)： 169-173. ［doi： 10.3969／j.issn.1674-8042.2018.02.011］
[full text view]
