基于RISC-V的身份证识别系统设计  

Design of ID card recognition system based on RISC-V

作　　者：张嘉 李新增 康鹏 金婕[1] ZHANG Jia;LI Xinzeng;KANG Peng;JIN Jie(College of Electronic and Electrical Engineering,Shanghai University of Engineering Science,Shanghai 201600,China)

机构地区：[1]上海工程技术大学电子电气工程学院,上海201600

出　　处：《传感器与微系统》2022年第8期89-91,95,共4页Transducer and Microsystem Technologies

摘　　要：设计了一种基于第5代精简指令系统(RISC-V)的SOC,用于身份证号码的在线识别,在Nexys4 DDR2开发板上验证并实现了整个系统。采用两种算法实现身份证号码识别,一种是TesserOCR识别引擎,一种是人工智能神经网络识别算法。实验结果表明:通过自制身份证号码数据集,在板上系统实现了98%的识别精度。A SOC based on the fifth-generation reduced instruction system RISC-V,which is used for online identification of ID numbers is designed,the entire system is verified and implemented on the Nexys4 DDR2 development board.Two algorithms are used to realize ID card number recognition,one is Tesserocr recognition engine,and the other is artificial intelligence neural network recognition algorithm.The experimental results show that through the self-made ID number dataset,98%recognition precision is achieved on the on-board system.

关 键 词：第5代精简指令系统 现场可编程门阵列 人工智能 

分 类 号：TP332[自动化与计算机技术—计算机系统结构]