正在加载图片...
工程科学学报.第42卷.第8期：1074-1084.2020年8月 Chinese Journal of Engineering,Vol.42,No.8:1074-1084,August 2020 https://doi.org/10.13374/j.issn2095-9389.2019.08.14.003;http://cje.ustb.edu.cn 弱光照条件下交通标志检测与识别 赵 坤，刘立，孟宇⑧，孙若灿 北京科技大学机械工程学院，北京100083 ☒通信作者.E-mail:myu@ustb.edu.cn 摘要针对弱光照条件下交通标志易发生漏检和定位不准的问题.本文提出了增强YOLOv3(You only look once)检测算 法，一种实时自适应图像增强与优化YOLO3网络结合的交通标志检测与识别方法.首先构建了大型复杂光照中国交通标 志数据集；然后针对复杂的弱光照图像提出自适应增强算法，通过调整图像亮度和对比度强化交通标志与背景之间的差异； 最后采用YOLO3网络框架检测交通标志.为了降低先验锚点框设置精度以及图像中背景与前景比例严重失衡对检测精度 造成的影响，优化了先验锚点框聚类算法和网络的损失函数.对比实验结果表明.在实时性大致相当的情况下，本文提出的 增强YOLOv3检测算法较标准YOLO3算法对交通标志有更高的回归精度和置信度，召回率和准确率分别提高0.96%和 0.48%. 关键词交通标志检测：弱光照：自适应图像增强；YOLOv3:深度学习 分类号TP391.4 Traffic signs detection and recognition under low-illumination conditions ZHAO Kun,LIU Li,MENG Yu,SUN Ruo-can School of Mechanical Engineering,University of Science and Technology Beijing,Beijing 100083,China Corresponding author,E-mail:myu@ustb.edu.cn ABSTRACT Traffic sign detection and recognition,which are important to ensure traffic safety,have been a research hotspot.In recent years,with the rapid development of automated driving technology,significant progress has been made in developing more accurate and efficient deep learning algorithms for traffic sign detection and recognition.However,these studies mainly focus on foreign traffic signs and do not consider the low-illumination conditions in practical application,which is a common scene.Therefore,many challenges still exist in the application of traffic sign detection and recognition in traffic scenes.To solve the problems of easy omission and inaccurate positioning for traffic sign detection and recognition under complex illumination conditions,the enhanced YOLOv3(You only look once)detection algorithm,a traffic sign detection and recognition method combining real-time adaptive image enhancement and the YOLOv3 frame was proposed.First,a large and complex illumination traffic sign dataset for Chinese traffic was constructed;it included globally low illumination,locally low illumination,and sufficient illumination images.Then an adaptive enhancement algorithm was proposed for low-illumination images,which can enhance the difference between traffic signs and background by adjusting the brightness and contrast of the images.Finally,high-quality and discrimination images as input were transmitted to the YOLOv3 network framework,and traffic sign detection and recognition were performed.To reduce the influences of the prior anchor box setting accuracy and the imbalance between the background and foreground on the detection accuracy,the clustering algorithm for the prior anchor box and loss function for the network were optimized.The results of the comparison experiment with the LISA dataset and complex illumination traffic sign dataset for Chinese traffic show that the proposed enhanced YOLOv3 detection algorithm has 收稿日期：2019-08-14 基金项目：国家重点研发计划资助项目(2018YFE0192900.2018YFC0810500.2018YFC0604403):国家高技术研究发展计划资助项目 (2011AA060408):中央高校基本科研业务费专项资金资助项目(FRF-TP.17-010A2)弱光照条件下交通标志检测与识别 赵 坤，刘 立，孟 宇苣，孙若灿 北京科技大学机械工程学院，北京 100083 苣通信作者，E-mail：myu@ustb.edu.cn 摘 要 针对弱光照条件下交通标志易发生漏检和定位不准的问题，本文提出了增强 YOLOv3（You only look once）检测算 法，一种实时自适应图像增强与优化 YOLOv3 网络结合的交通标志检测与识别方法. 首先构建了大型复杂光照中国交通标 志数据集；然后针对复杂的弱光照图像提出自适应增强算法，通过调整图像亮度和对比度强化交通标志与背景之间的差异； 最后采用 YOLOv3 网络框架检测交通标志. 为了降低先验锚点框设置精度以及图像中背景与前景比例严重失衡对检测精度 造成的影响，优化了先验锚点框聚类算法和网络的损失函数. 对比实验结果表明，在实时性大致相当的情况下，本文提出的 增强 YOLOv3 检测算法较标准 YOLOv3 算法对交通标志有更高的回归精度和置信度，召回率和准确率分别提高 0.96% 和 0.48%. 关键词 交通标志检测；弱光照；自适应图像增强；YOLOv3；深度学习 分类号 TP391.4 Traffic signs detection and recognition under low-illumination conditions ZHAO Kun，LIU Li，MENG Yu苣 ，SUN Ruo-can School of Mechanical Engineering, University of Science and Technology Beijing, Beijing 100083, China 苣 Corresponding author, E-mail: myu@ustb.edu.cn ABSTRACT Traffic sign detection and recognition, which are important to ensure traffic safety, have been a research hotspot. In recent years, with the rapid development of automated driving technology, significant progress has been made in developing more accurate and efficient deep learning algorithms for traffic sign detection and recognition. However, these studies mainly focus on foreign traffic signs and do not consider the low-illumination conditions in practical application, which is a common scene. Therefore, many challenges still exist in the application of traffic sign detection and recognition in traffic scenes. To solve the problems of easy omission and inaccurate positioning for traffic sign detection and recognition under complex illumination conditions, the enhanced YOLOv3 (You only look once) detection algorithm, a traffic sign detection and recognition method combining real-time adaptive image enhancement and the YOLOv3 frame was proposed. First, a large and complex illumination traffic sign dataset for Chinese traffic was constructed; it included globally low illumination, locally low illumination, and sufficient illumination images. Then an adaptive enhancement algorithm was proposed for low-illumination images, which can enhance the difference between traffic signs and background by adjusting the brightness and contrast of the images. Finally, high-quality and discrimination images as input were transmitted to the YOLOv3 network framework, and traffic sign detection and recognition were performed. To reduce the influences of the prior anchor box setting accuracy and the imbalance between the background and foreground on the detection accuracy, the clustering algorithm for the prior anchor box and loss function for the network were optimized. The results of the comparison experiment with the LISA dataset and complex illumination traffic sign dataset for Chinese traffic show that the proposed enhanced YOLOv3 detection algorithm has 收稿日期: 2019−08−14 基金项目: 国家重点研发计划资助项目（ 2018YFE0192900， 2018YFC0810500， 2018YFC0604403） ；国家高技术研究发展计划资助项目 （2011AA060408）；中央高校基本科研业务费专项资金资助项目（FRF-TP-17-010A2） 工程科学学报，第 42 卷，第 8 期：1074−1084，2020 年 8 月 Chinese Journal of Engineering, Vol. 42, No. 8: 1074−1084, August 2020 https://doi.org/10.13374/j.issn2095-9389.2019.08.14.003; http://cje.ustb.edu.cn