基于深度学习的雾天交通标志检测系统设计  

Design of Foggy Traffic Sign Detection System Based on Deep Learning

作　　者：张慧[1] ZHANG Hui(QuFu Normal University Library,Qufu Shandong 273100,China)

机构地区：[1]曲阜师范大学图书馆,山东曲阜273100

出　　处：《信息与电脑》2024年第18期100-102,共3页Information & Computer

摘　　要：针对雾天环境能见度低、采集的图像质量差导致的交通标志检测准确率低、易出现漏检、误检等问题,本文设计并实现了一款基于深度学习的雾天交通标志检测系统。首先,采用AOD-NET算法对采集到的雾天图像进行去雾处理,然后将YOLOv5算法作为目标检测算法,对去雾后的图像进行交通标志进行检测与识别,从而实现了雾天环境下的交通标志检测。测试结果表明,该系统能够有效地提高雾天天气情况下交通标志的检测准确率,减少误检、漏检,实现了交通标志快速准确的检测和识别。Aiming at the problems of low visibility and poor image quality in foggy environment,which lead to low accuracy and false detections of traffic sign,a traffic sign detection system in foggy based on deep learning was designed and implemented in this paper.Firstly,the AOD-NET algorithm is used to remove fog from the collected foggy images.Secondly,YOLOv5 algorithm is adopted as the target detection algorithm to realize the traffic sign detection in foggy environment.The test results show that this system can effectively improve the detection accuracy of traffic signs in foggy environment,reduce false positives and omissions,and achieve fast and accurate detection and recognition of traffic signs.

关 键 词：交通标志识别 雾天图像 YOLOv5 AOD-Net 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]