摘要
设计并实现了一种基于交通标识识别的自动巡线智能车,通过搭建神经网络的图像处理模型和车载图像处理技术来识别路径上的交通标识和路径信息,得出识别的结果并做出相应控制。详细介绍了智能车的系统硬件组成、深度学习模型的建立、所需数据的采集方法以及模型训练的方法、实验结果与分析。结果表明,设计能够满足小车自动巡线过程中的交通标识识别,并满足了便携性、准确性、快速性等要求。
This paper designs and implements an automatic patrol intelligent vehicle based on traffic sign recognition,which identifies traffic signs and path information by building an image processing model of neural network and on-board image processing technology,derives the recognition results and makes corresponding control.The system hardware composition of the intelligent vehicle,the establishment of the deep learning model,the method of collecting the required data and the method of model training,the experimental results and analysis are described in detail.The results show that the design can meet the traffic sign recognition in the automatic cart patrol process and satisfy the requirements of portability,accuracy and speed.
作者
徐鹏飞
周燕云
刁鹏程
夏乐天
XU Peng-fei;ZHOU Yan-yun;DIAO Peng-cheng;XIA Le-tian(School of Information Science and Engineering,Xinjiang University,Xinjiang Urumqi,830046,China)
出处
《辽宁工业大学学报(自然科学版)》
2023年第2期108-112,共5页
Journal of Liaoning University of Technology(Natural Science Edition)
基金
新疆大学实验室管理与实践研究项目(2023年)。
关键词
深度学习
交通标识识别
树莓派
自动巡线
deep learning
traffic sign recognition
raspberry Pi
automatic patrol intelligent vehicle