[1]宋佳蓉,杨忠,张天翼,等.基于卷积神经网络和多类SVM的交通标志识别[J].应用科技,2018,45(05):71-75,81.[doi:10.11991/yykj.201710009]
 SONG Jiarong,YANG Zhong,ZHANG Tianyi,et al.Traffic signs recogniton based on convolutional neural networks and multi-class SVM[J].Applied science and technology,2018,45(05):71-75,81.[doi:10.11991/yykj.201710009]
点击复制

基于卷积神经网络和多类SVM的交通标志识别(/HTML)
分享到:

《应用科技》[ISSN:1009-671X/CN:23-1191/U]

卷:
第45卷
期数:
2018年05期
页码:
71-75,81
栏目:
自动化技术
出版日期:
2018-09-15

文章信息/Info

Title:
Traffic signs recogniton based on convolutional neural networks and multi-class SVM
作者:
宋佳蓉 杨忠 张天翼 韩家明 朱家远
南京航空航天大学 自动化学院, 江苏 南京 211106
Author(s):
SONG Jiarong YANG Zhong ZHANG Tianyi HAN Jiaming ZHU Jiayuan
College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China
关键词:
交通标志识别卷积神经网络深度学习迁移学习多类SVM过拟合softmaxAlexNet
Keywords:
traffic signs recognitionCNNsdeep learningtransfer learningmulti-class SVMoverfittingsoftmaxAlexNet
分类号:
TP751
DOI:
10.11991/yykj.201710009
文献标志码:
A
摘要:
为了实现在复杂环境下具有较高准确率的交通标志识别以及在小样本情况下也能良好工作的识别网络,提出一种基于卷积神经网络和多类SVM的交通标志识别模型。此模型不需人工设计特征提取算法,且在小样本训练集上也能训练出具有较高准确率的分类模型。除此之外,利用迁移学习策略,避免重新初始化卷积神经网络,在节省大量样本与训练时间的同时能有效避免过拟合的发生。实验结果表明,提出的分类模型在小样本训练集上进行训练后得到的模型在实际测试中有较好的表现且对处于复杂背景下和严重畸变的交通标志具有可靠的识别能力和良好分类结果。
Abstract:
In order to realize recognition of traffic signs with higher accuracy in complex environment and to acquire a classifying network that works well with small samples, a traffic signs recognition system based on the convolutional neural networks (CNNs) and multi-class support vector machine (SVM) is proposed in this paper. It does not need to design artificial feature extraction algorithm and a classification model with higher accuracy can be trained on a small sample training set. Besides, the training learning strategy is used to avoid reinitializing CNNs, so it avoids overfitting while saving a large amount of samples and time efficiently. Experimental results demonstrate that the classification model obtained in the small sample training set has good performance and has reliable recognition ability and good classification results for samples even in the condition of complex background and severe distortion.

参考文献/References:

[1] 王方石, 王坚, 李兵, 等. 基于深度属性学习的交通标志检测[J]. 吉林大学学报:工学版, 2018, 48(1):319-329
[2] 吕耀坤. 基于卷积神经网络的实景交通标志识别[J]. 物联网技术, 2017, 7(1):29-30
[3] 黄琳, 张尤赛. 应用深层卷积神经网络的交通标志识别[J]. 现代电子技术, 2015, 38(13):101-106
[4] 郭显娥, 武伟, 刘春贵, 等. 多类SVM分类算法的研究[J]. 山西大同大学学报:自然科学版, 2010, 26(3):6-8, 14
[5] TAJBAKHSH N, SHIN J Y, GURUDU S R, et al. Convolutional neural networks for medical image analysis:full training or fine tuning[J]. IEEE transactions on medical imaging, 2016, 35(5):1299-1312.
[6] MARMANIS D, DATCU M, ESCH T, et al. Deep learning earth observation classification using ImageNet Pretrained networks[J]. IEEE geoscience and remote sensing letters, 2016, 13(1):105-109.
[7] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[C]//Proceedings of the 25th International Conference on Neural Information Processing Systems. Lake Tahoe, Nevada, 2012:1097-1105.
[8] GU Yingxin, WYLIE B K, BOYTE S P, et al. An optimal sample data usage strategy to minimize overfitting and underfitting effects in regression tree models based on remotely-sensed data[J]. Remote sensing, 2016, 8(11):943.
[9] SRIVASTAVA N, HINTON G, KRIZHEVSKY A, et al. Dropout:a simple way to prevent neural networks from overfitting[J]. Journal of machine learning research, 2014, 15(1):1929-1958.
[10] 张冬生. 支持向量机在分类问题中的应用研究[J]. 黑龙江科技信息, 2010(35):64
[11] WU Yichao, LIU Yufeng. Robust truncated hinge loss support vector machines[J]. Journal of the American statistical association, 2007, 102(479):974-983.
[12] 刘建伟, 李双成, 罗雄麟. p范数正则化支持向量机分类算法[J]. 自动化学报, 2012, 38(1):76-87

相似文献/References:

[1]郜雨桐,宁慧,王巍,等.基于卷积神经网络的车辆型号识别研究[J].应用科技,2018,45(06):53.[doi:10.11991/yykj.201803011]
 GAO Yutong,NING Hui,WANG Wei,et al.Research on vehicle model identification based on convolutional neural network[J].Applied science and technology,2018,45(05):53.[doi:10.11991/yykj.201803011]
[2]王立新,江加和.基于深度学习的显著性区域的图像检索研究[J].应用科技,2018,45(06):63.[doi:10.11991/yykj.201803012]
 WANG Lixin,JIANG Jiahe.Research on image retrieval of saliency region based on deep learning[J].Applied science and technology,2018,45(05):63.[doi:10.11991/yykj.201803012]

备注/Memo

备注/Memo:
收稿日期:2017-10-29。
基金项目:国家自然科学基金项目(61473144);航空科学基金项目(20162852031);科技部重大科学仪器设备开发专项子课题(2016YFF0103702)
作者简介:宋佳蓉(1993-),女,硕士研究生;杨忠(1969-),男,教授,博士生导师
通讯作者:杨忠,E-mail:YangZhong@nuaa.edu.cn
更新日期/Last Update: 2018-09-04