摘要
提出了一种基于红外光源和低成本黑白CCD摄像机的,以TMS320DM642为核心信号处理器的虹膜采集识别系统。采用红外光源解决了照明问题,使图像更清晰。外扩的SDRAM和FLASH使得图像存储问题得以解决。在虹膜识别算法的实现上,通过使用多方向2D-Gabor滤波器组提取虹膜局部方向特征并形成简单易处理的二进制形式的方向特征码,通过改进的Hamming距离进行匹配。实验证明该采集识别系统识别准确率高,速度快。
This paper proposes an iris sample and recognition system based on infrared light source, low cost black and white CCD camera and TMS320DM642 as its core DSP. The system takes advantage of the infrared light source to solve the lighting problem so that pictures sampled are clearer and uses the extent SDRAM and FLASH to settle the storage problem. As the recognition algorithm, it uses multi-direction 2D-Gabor filters group to extract iris local direction features, codes them into a binary code that is easy to deal and matches are achieved by the improved Hamming distance. Experiments show that this sample and recognition system wins a high speed and recognition rate.
出处
《微计算机信息》
2009年第29期21-23,共3页
Control & Automation
基金
基金申请人:苑玮琦
项目名称:人眼自然张开状态下虹膜识别方法的研究
虹膜识别几个关键问题的研究
基金颁发部门:国家自然科学基金(60672078
60472088)
