| 基于深度学习的交通标志检测系统仿真 |
| |
| 引用本文： | 李克俭,陈少波,李万琦.基于深度学习的交通标志检测系统仿真[J].教育技术导刊,2009,19(9):31-34. |
| |
| 作者姓名： | 李克俭 陈少波 李万琦 |
| |
| 作者单位： | 中南民族大学 电子信息工程学院，湖北 武汉 430074 |
| |
| 基金项目： | 中南民族大学中央高校基本科研业务费专项资金项目（CZY18002） |
| |
| 摘 要： | 交通标志是规范驾驶员驾驶的重要指标信息，如何检测交通标志是无人驾驶和辅助驾驶中的关键一环。利用 PYQT 开发一套基于深度学习的交通标志检测系统，系统包括 4 个主要模块：用户信息模块、摄像头采集模块、检测模块和保存结果模块。对比目前主流的深度学习目标检测算法 YOLOv3 和 Faster-RCNN 在交通标志上的检测效果，并采用 YOLOv3 作为系统仿真算法，仿真结果表明，YOLOv3 兼顾了实时检测和检测精度要求，对无人驾驶和辅助驾驶研究具有一定应用价值。
|
| 关 键 词： | 深度学习 目标检测 交通标志 YOLO Faster-RCNN |
| 收稿时间： | 2020-01-13 |
Simulation of Traffic Sign Detection System Based on Deep Learning |
| |
| LI Ke-jian,CHEN Shao-bo,LI Wan-qi.Simulation of Traffic Sign Detection System Based on Deep Learning[J].Introduction of Educational Technology,2009,19(9):31-34. |
| |
| Authors: | LI Ke-jian CHEN Shao-bo LI Wan-qi |
| |
| Institution: | School of Electronics and Information Engineering，South-Central University for Nationalities，Wuhan 430074，China |
| |
| Abstract: | Traffic identifier is an important index information for regulating drivers. How to detect traffic identifiers is a key part of driverless and assisted driving. This paper uses PYQT to develop a traffic sign detection system for deep learning. The system includes four main modules：a user information module，a camera acquisition module，a detection module and a save result module. The current mainstream deep learning algorithms YOLO and Faster-RCNN in the detection of traffic identifiers are compared. In the end，YOLO is used as the algorithm for system simulation. The system simulation results prove that YOLO can achieve both real-time monitoring and accuracy requirements in terms of detection accuracy and real-time performance，and has certain application value for the research on unmanned driving and assisted driving. |
| |
| Keywords: | deep learning object detection traffic sign YOLO Faster R-CNN |
|
| 点击此处可从《教育技术导刊》浏览原始摘要信息 |
| 点击此处可从《教育技术导刊》下载免费的PDF全文 |
