张邯,罗晓曙,袁荣尚

• 1:广西师范大学电子工程学院

摘要(Abstract)：

在现实的交通环境中,由于各种因素影响,使得所采集到的交通标志图像识别的准确性不高,鲁棒性也较差,给交通标志的准确识别带来了很大的困难。为此,采用非对称卷积结构对经典卷积神经网络AlexNet进行改进,并引入批量归一化(BN)方法,提出基于优化卷积神经网络结构的交通标志识别方法。其中,非对称卷积结构使网络进一步加深,提高了识别精度。BN将每一层的输出数据归一化为均值为0、标准差为1,确保了数据稳定,使梯度传输更为顺畅。使用德国交通标志数据集进行训练并测试,结果显示改进的网络结构提升了网络的分类精度,且达到了97.56%,具有一定的应用价值。

关键词(KeyWords)： 卷积神经网络;非对称卷积;批量归一化;交通标志;梯度传输;分类精度

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金(11262004);; 广西多源信息挖掘与安全重点实验室开放基金(MIMS15-06);; 广西信息科学实验中心基金(KA1430);; 广西研究生教育创新计划项目(XYCSZ2017051)~~

作者(Author): 张邯,罗晓曙,袁荣尚

DOI: 10.16652/j.issn.1004-373x.2018.21.030

参考文献(References)：

• [1]刘平华,李建民,胡晓林,等.动态场景下的交通标识检测与识别研究进展[J].中国图象图形学报,2013,18(5):493-503.LIU Pinghua,LI Jianmin,HU Xiaolin,et al. Recent progress in detection and recognition of the traffic signs in dynamic scenes[J]. Journal of image and graphics,2013,18(5):493-503.
• [2] SAHA S K,DULAL C M,BHUIYAN A A. Neural network based sign recognition[J]. International journal of computer application,2012,50(10):35-41.
• [3] HOWARD A G. Some improvements on deep convolutional neural network based image classification[J]. Computer science,2013,4(2):1-6.
• [4]柯小龙.卷积神经网络图像分类应用研究[D].深圳:深圳大学,2017.KE Xiaolong. Applied research of CNN network image[D].Shenzhen:Shenzhou University,2017.
• [5] 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:Curran Associates Inc.,2012:1097-1105.
• [6] HOCHREITER S. The vanishing gradient problem during learning recurrent neural nets,and problem solutions[J]. International journal of uncertainty,fuzziness and knowledge-based systems,1998,6(2):107-116.
• [7] HINTON G E,SRIVASTAVA N,KRIZHEVSKY A,et al. Improving neural networks by preventing co-adaptation of feature detectors[J]. Computer science,2012,3(4):212-223.
• [8] LIN M,CHEN Q,YAN S. Network in network[J]. Computer science,2013,4(5):1-10.
• [9] SZEGEDY C,LIU W,JIA Y,et al. Going deeper with convolutions[C]//2015 IEEE Conference on Computer Vision and Pattern Recognition. Boston:IEEE,2015:1-9.
• [10] SIMONYAN K,ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[J]. Computer science,2014,5(2):1-7.
• [11] HE K,ZHANG X,REN S,et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas:IEEE,2016:770-778.
• [12]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.HUANG Lin,ZHANG Yousai. Traffic signs recognition applying with deep-layer convolution neural network[J]. Modern electronics technique,2015,38(13):101-106.
• [13] RUMELHART D E,HINTON G E,WILLIAMS R J. Learning internal representation by back-propagation of errors[J].Nature,1986,323:533-536.
• [14] SZEGEDY C,VANHOUCKE V,IOFFE S,et al. Rethinking the inception architecture for computer vision[C]//2016 IEEE Conference on Computer vision and pattern recognition. Las Vegas:IEEE,2015:2818-2826.
• [15]杨天祺,黄双喜.改进卷积神经网络在分类与推荐中的实例应用[J].计算机应用研究,2018,35(4):974-977.YANG Tianqi,HUANG Shuangxi. Application of improved convolution neural network in classification and recommendation[J]. Application research of computers,2018,35(4):974-977.
• [16] JIA Yangqing,SHELHAMER E,DONAHUE J,et al. Caffe:convolutional architecture for fast feature embedding[J]. Computer science,2014,5(7):675-678.
• [17] HOUBEN S,STALLKAMP J,SALMEN J,et al. Detection of traffic signs in real-world images:the German traffic sign detection benchmark[C]//Proceedings of the 2013 International Joint Conference on Neural Networks. Washington, DC:IEEE,2013:715-722.