摘要
近几年,随着智能交通系统的发展,我国汽车保有量迅速增长,套牌车和无牌车大量出现,违法行为频繁发生.如何能快速准确地识别车辆标志,已成为交通信息安全管理部门亟待解决的问题.为了能快速准确地识别车辆标志,首先将忆阻器和卷积神经网络结合起来进行车辆标志识别,提出一种全新的数字图像预处理算法.为了模拟真实场景中的噪声,基于VLR-40数据集构造4类新的数据集.然后通过2大实验对基于忆阻神经网络的车辆标志识别技术进行深入研究与分析.最后对车辆标志识别技术在智能交通信息安全中的应用进行了探讨.
In recent years,with the development of intelligent transportation systems,the number of cars in our country has grown rapidly.However,with a large number of deck cars and unlicensed cars appearing,and illegal acts frequently occur.How to identify vehicle logo quickly and accurately has become an urgent problem for traffic management departments.In order to quickly and accurately identify the vehicle,this dissertation first proposes a new digital image preprocessing algorithm by combining the memristor and convolutional neural network for vehicle logo recognition.In order to simulate the noise in the real scene,four new datasets are constructed based on the VLR-40 dataset.Through two experiments,the vehicle logo recognition technology based on memristive neural network is deeply researched and analyzed.Finally,the application of vehicle logo recognition technology in intelligent traffic information security is discussed.
作者
罗运鑫
佘堃
于钥
李洋
Luo Yunxin;She Kun;Yu Yue;Li Yang(School of Information and Software Engineering,University of Electronic Science and Technology of China,Chengdu 610054;Open Project Fund of Intelligent Terminal Key Laboratory of Sichuan Province(Yibin Institute of Universityof Electronic Science and Technology of China),Yibin,Sichuan 644002)
出处
《信息安全研究》
2021年第8期715-727,共13页
Journal of Information Security Research
基金
厅市共建智能终端四川省重点实验室开放基金项目(2019—2020)(SCITLAB-0002)
四川省重大科技专项(2018GZDZX0012)。
