基于FPGA和BNN的高效车牌识别系统设计  

Efficient license plate recognition system based on FPGA and BNN

作　　者：袁敏 张振东 YUAN Min;ZHANG Zhendong(School of Artificial Intelligence,University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院大学人工智能学院,北京100049

出　　处：《集成电路与嵌入式系统》2024年第10期19-24,共6页Integrated Circuits and Embedded Systems

摘　　要：针对传统车牌识别系统处理速度慢、准确率低、硬件资源消耗高等问题,设计并实现了一种高效的车牌识别系统。该系统基于现场可编程门阵列(FPGA)与二值神经网络(BNN)技术,通过结合硬件加速与算法优化,显著提高了车牌识别性能。实验结果表明,系统的识别准确率达到96.46%,识别时间缩短至12 ms。与传统车牌识别算法和CNN FPGA方案相比,该系统在硬件资源消耗、识别速度和准确率方面表现出明显优势,为高效、资源友好的车牌识别提供了有效解决方案。To address the issues of slow processing speed,low accuracy,and high hardware resource consumption in traditional license plate recognition systems,an efficient license plate recognition system is designed and implemented.This system is based on Field Programmable Gate Array(FPGA)and Binary Neural Network(BNN)technology.By combining hardware acceleration with algorithm optimization,the system significantly enhances license plate recognition performance.The experimental results show that the system achieves a recognition accuracy of 96.46%,reducing the recognition time to 12 milliseconds.Compared to traditional license plate recognition algorithms and CNN FPGA solutions,this system demonstrates significant advantages in hardware resource consumption,recognition speed,and accuracy,providing an efficient and resource-friendly solution for license plate recognition.

关 键 词：FPGA BNN 车牌识别系统 硬件加速 算法优化 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]