基于Zynq的虹膜识别系统设计与实现  

Design and implementation of iris recognition system based on Zynq

作　　者：李萌 徐驰[1] 俞度立 LI Meng;XU Chi;YU Duli(Beijing University of Chemical Technology,Beijing 100029,China)

机构地区：[1]北京化工大学信息与技术学院,北京100029 [2]北京化工大学软物质科学与工程高精尖创新中心,北京100029

出　　处：《自动化与仪器仪表》2020年第2期8-11,共4页Automation & Instrumentation

基　　金：北京化工大学高精尖创新团队项目(No.buctylkjcx06)

摘　　要：针对虹膜识别系统存在成本高、操作难度大等不足,给出一种基于Zynq的虹膜识别系统设计方法,进一步拓展虹膜识别系统在民用市场中的运用面。具体讲,就是将虹膜图像存储在SD卡内,通过linux的ramfs(am file system)将SD卡中的虹膜图像加载到RAM,使用VDMA将虹膜数据从RAM搬运至虹膜降噪IP核进行算法加速,完成矩阵卷积的硬件加速,加速完成的数据通过VDMA,搬运回至RAM,供虹膜处理的下一个阶段使用。实验结是证明了所提方法的有效性。With respect to the high cost and difficulty of operation of the iris recognition system,A design method of iris recognition system based on Zynq is presented,and the application of iris recognition system is further expanded in the civilian market.Specifically,the iris image is stored in the SD card,the iris image in the SD card is loaded into the RAM through the ramfs(ram file system)of linux,and the iris data is transferred from the RAM to the iris noise reduction IP core using VDMA,and the iris matrix is entered into the iris noise reduction IP core to accelerate the algorithm,and carry it back to RAM for the next stage of iris processing.The experimental results demonstrate the effectiveness of the proposed method.

关 键 词：虹膜识别 图像存储 降噪IP核 算法加速 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]