目标检测与模板匹配相结合的交通标志小目标识别算法研究  

Research on traffic sign small target recognition algorithm combining object detection and template matching

作　　者：张旭 郑清春[1,2,3,4] 赵阳阳 朱培浩[3] ZHANG Xu;ZHENG Qingchun;ZHAO Yangyang;ZHU Peihao(School of Computer Science and Engineering,Tianjin University of Technology,Tianjin 300384,China;Key Laboratory of Computer Vision and System,Ministry of Education,Tianjin University of Technology,Tianjin 300384,China;School of Mechanical Engineering,Tianjin University of Technology,Tianjin 300384,China;Tianjin Vocational Institute,Tianjin 300410,China)

机构地区：[1]天津理工大学计算机科学与工程学院,天津300384 [2]天津理工大学计算机视觉与系统教育部重点实验室,天津300384 [3]天津理工大学机械工程学院,天津300384 [4]天津职业大学,天津300410

出　　处：《智能计算机与应用》2025年第3期186-191,共6页Intelligent Computer and Applications

基　　金：天津市研究生科研创新项目(2021YJSB242)。

摘　　要：交通标志识别作为交通和路况的提示信息,是智能驾驶的重要一环。通常都是单纯采用目标检测算法进行交通标志识别,忽略了交通标志本身所具有的标准化和单一性。本文基于交通标志的标准化和个性化,提出一种更有效的交通标志识别算法。首先对自然场景下的交通标志数据集TT100K中的交通标志增加所属的大类信息,然后利用目标检测算法检测数据集中的标志所属大类,获得目标检测框,最后将标准交通标志图片作为模板,通过模板匹配方法与目标检测框区域内图像进行相似性对比,确定交通标志具体种类。实验证明,本文提出的算法与单纯使用目标检测算法相比大大提高了识别精度,为交通标志小目标识别算法的研究提供了新的思路。Traffic sign recognition,as a reminder of traffic and road conditions,is an important part of intelligent driving.Usually,for traffic sign recognition,only target detection algorithms are used,ignoring the standardization and singularity of traffic signs themselves.This paper proposes a more effective traffic sign recognition algorithm based on the standardization and personalization of traffic signs.Firstly,the category information of traffic signs in the natural scene traffic sign dataset TT100K is added,and then the object detection algorithm is used to detect the category of signs in the dataset to obtain the target detection box.Finally,standard traffic sign images are used as templates,and the similarity between the template matching method and the images in the object detection box area is compared to determine the specific types of traffic signs.Experimental results have shown that the algorithm proposed in this paper greatly improves recognition accuracy compared to the simple use of object detection algorithms,providing new ideas for the research of small target recognition algorithms for traffic signs.

关 键 词：目标检测 模板匹配 交通标志 小目标识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]