Please wait a minute...
浙江大学学报(工学版)
浙江大学学报(工学版)  2018, Vol. 52 Issue (4): 694-702    DOI: 10.3785/j.issn.1008-973X.2018.04.012
自动化技术     
基于深度卷积神经网络的车型识别方法
袁公萍, 汤一平, 韩旺明, 陈麒
浙江工业大学 信息工程学院, 浙江 杭州 310023
Vehicle category recognition based on deep convolutional neural network
YUAN Gong-ping, TANG Yi-ping, HAN Wang-ming, CHEN Qi
School of Information Engineering, Zhejiang University of Technology, Hangzhou 310023, China
 全文: PDF(4177 KB)   HTML
摘要：

针对现有车辆车型视觉识别技术中的检测精度不高、难以适应天气环境变化、难以从视频图像中准确提取出用于识别的车辆图像、难以对车辆车型子类进行识别分类、难以兼顾识别精度和检测效率等不足，将深度卷积神经网络引入车辆目标定位、识别和分类（子类）问题中.利用深度卷积神经网络自动完成车型的深度特征学习，在特征图上进行逻辑回归，从道路复杂背景中提取出感兴趣区域；利用softmax分类器训练特征实现车型识别；为了优化softmax在深度卷积神经网络分类过程中出现的类内间距大的问题，引入中心损失函数对softmax损失函数进行优化，提高类间分散性与类内紧密性.在BIT-Vehicle车型数据集中的实验结果显示，提出方法的平均精度为89.67%，检测和识别时间为159 ms；与传统的分类方法相比，识别精度提高约20%，效率提高10倍以上，检测鲁棒性有明显提升；与未改进前的深度卷积神经网络相比，检测精度提高0.6%，速度提高0.29倍.
Abstract:

A novel detection model was proposed to solve the problems of low detection accuracy, difficulty in adapting to changeable climate environment, extracting the vehicles from the video accurately, meeting fine classification based on visual method, and giving attention to both the recognition accuracy and the detection efficiency and so on in the existing vehicle identification and classification. The problems contain three consecutive stages:vehicle detection, features extraction and classification based on deep convolutional neural network. The deep learning network was used to automatically complete vehicle feature extraction and logical regression was performed on the feature map to extract the region of interest from the complex background. Then vehicle recognition was implemented to train the existing features through a softmax classifier. The central loss function was introduced to optimize this problem and improve inter-class dispension and intra-class compactness as much as possible in order to optimize the problem that softmax will cause large class spacing in deep convolutional neural network. Experiments on BIT-Vehicle dataset demonstrated that the identification average accuracy of our algorithm was about 89.67%, and the recognition rate was about 159 ms. The recognition accuracy increased by approximately 20% compared with conventional machine learning methods, and the rate increased by at least 10 times. Robustness was significantly improved. The recognition accuracy increased by approximately 0.6% and the rate increased by at least 0.29 times compared with the unimproved deep convolution neural network.
收稿日期: 2017-01-12
CLC:  TP391  
基金资助:

国家自然科学基金资助项目（61070134，61379078）.
通讯作者: 汤一平,男,教授.orcid.org/0000-0002-7128-2784.     E-mail: typ@zjut.edu.cn
作者简介: 袁公萍(1992-),男,硕士,从事计算机视觉与深度学习研究.orcid.org/0000-0002-6489-7676.E-mail:1030617785@qq.com
服务　　
把本文推荐给朋友
加入引用管理器
E-mail Alert
作者相关文章　　

引用本文:

袁公萍, 汤一平, 韩旺明, 陈麒. 基于深度卷积神经网络的车型识别方法[J]. 浙江大学学报(工学版), 2018, 52(4): 694-702.

YUAN Gong-ping, TANG Yi-ping, HAN Wang-ming, CHEN Qi. Vehicle category recognition based on deep convolutional neural network. JOURNAL OF ZHEJIANG UNIVERSITY (ENGINEERING SCIENCE), 2018, 52(4): 694-702.

链接本文:

http://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2018.04.012    或    http://www.zjujournals.com/eng/CN/Y2018/V52/I4/694

[1] YIN Q, CAO Z M, JIANG Y N, et al. Learning deep face representation:U.S. Patent 20150347820[P]. 2015-12-03.
[2] WEI W, ZHANG Q S, WANG M J. A method of vehicle classification using models and neural networks[C]//IEEE 2001 International Conference on Semiannual Vehicular Technology 2001 Spring Conference (VTC). Israel:IEEE, 2001:3022-3026.
[3] LI L Y, HUANG W M, GU I Y H, et al. Foreground object detection from videos containing complex background[C]//11th ACM International Conference on Multimedia. Berkeley:ACM, 2003:2-10.
[4] 张旭东, 钱玮, 高隽, 等. 基于稀疏贝叶斯分类器的汽车车型识别[J].小型微型计算机系统, 2005, 26(10):1839-2841. ZHANG Xu-dong, QIAN Wei, GAO Jun, et al. Vehicle model recognition system based on sparse Bayesian classification[J]. Journal of Chinese Computer Systems.2005, 26(10):1839-2841.
[5] XIA L M. Vehicle recognition using boosting neural network classifiers[C]//Proceedings of the 6th World Congress on Intelligent Control and Automation (ICA). Dalian:[s. n.], 2006:9641-9644.
[6] SIDDIQUE M N H, TOKHI M O. Training neural networks:backpropagation VS. genetic algorithms[C]//International Joint Conference on Neural Networks. Washington, DC:IEEE, 2001:2673-2678.
[7] 王守觉.仿生模式识别(拓扑模式识别)一种模式识别新模型的理论与应用[J].电子学报, 2002, 30(10):1417-1420. WANG Shou-jue. Bionic (topological) pattern recognition a new model of pattern recognition theory and its applications[J]. Acta Electronica Sinica,2002, 30(10):1417-1420.
[8] 陈爱斌. 基于支持向量机的车型识别[D].湖南:中南大学, 2004:32. CHEN Ai-bin. Vehicle identification based on support vector machine[D].Hunan:Central South University,2014:32.
[9] YANG K H, SHAN G L, ZHAO L L. Application of least squares support vector machine on vehicle recognition[C]//International Conference on Intelligent Systems Design and Applications (ISDA). Jinan:IEEE, 2006:217-221.
[10] 武宏伟, 马钱. 一种基于支持向量机的车型自动分类器设计方案[J].计算机应用, 2005(4):10-12. WU Hong-wei, MA Qian. A design scheme of automatic vehicle classification based on support vector machine[J]. Journal of Computer Application, 2005(4):10-12.
[11] MA X, GRIMSON W E L. Edge-based rich representation for vehicle classification vision[C]//10th IEEE International Conference on Computer Vision (ICCV). Beijing:IEEE, 2005:1185-1192.
[12] GE F X, SHI Y, SUN B,et al. Sparse representation based classification by using PCA-SIFT descriptors[C]//IEEE International Conference on Information Science and Technology (ICIST).Guangdong:IEEE, 2014:429-432.
[13] HAN F,SHAN Y,CEKANDER R,et al. A two-stage approach to people and vehicle detection with hog-based SVM[C]//Performance Metrics for Intelligent Systems Workshop in conjunction with the IEEE Safety, Security, and Rescue Robotics Conference (PerMIS, SSRR). Washington, D.C:IEEE, 2006:133-140.
[14] GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Columbus:IEEE, 2014:580-587.
[15] GIRSHICK R. Fast R-CNN[C]//Proceedings of the IEEE International Conference on Computer Vision (ICCV). Boston:IEEE, 2015:1440-1448.
[16] REN S Q, HE K M, GIRSHICK R, et al. Faster rcnn:towards real-time object detection with region proposal networks[C]//Advances in Neural Information Processing Systems (NIPS). Montreal:[s. n.], 2015:91-99.
[17] HE K M, ZHANG X Y, REN S Q, et al. Deep Residual Learning for Image Recognition[C]//Computer Vision and Pattern Recognition (CVPR). Las Vegas:[s. n.], 2016:770-778.
[18] WEN Y D, ZHANG K P, LI Z F, et al. A discriminative feature learning approach for deep face recognition[C]//European Conference on Computer Vision (ECCV). Amsterdam:[s. n.], 2016:499-515.
[19] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[C]//International Conference on Neural Information Processing Systems (NIPS). Lake Tahoe:[s. n.], 2012:1097-1105.
[20] ZEILER M D, FERGUS R. Visualizing and understanding convolutional neural networks[C]//European Conference on Computer Vision. Springer International Publishing (ECCV). Zurich:[s. n.], 2014:818-833.
[21] LECUN Y, CORTES C, BURGER C J. The mnist database of handwritten digits. 1998-11. http://yann.lecun.com/exdb/mnist/.
[22] ZHEN D, WU Y W, PEI M T, et al. Vehicle type classification using a semisupervised convolutional neural network[J]. IEEE Transactions on Intelligent Transportation Systems, 2015, 16(4):2247-2256.
[23] UIJLINGS J R R, SANDE K E A V D, GEVERS T, et al. Selective search for object recognition[J]. International Journal of Computer Vision, 2013, 104(2):154-171.
[24] DOLLAR P, ZITNICK C L. Structured forests for fast edge detection[C]//IEEE International Conference on Computer Vision (ICCV). Sydney:IEEE, 2013:1841-1848.
[1] 韩勇, 宁连举, 郑小林, 林炜华, 孙中原. 基于社交信息和物品曝光度的矩阵分解推荐[J]. 浙江大学学报(工学版), 2019, 53(1): 89-98.
[2] 郑洲, 张学昌, 郑四鸣, 施岳定. 基于区域增长与统一化水平集的CT肝脏图像分割[J]. 浙江大学学报(工学版), 2018, 52(12): 2382-2396.
[3] 赵丽科, 郑顺义, 王晓南, 黄霞. 单目序列的刚体目标位姿测量[J]. 浙江大学学报(工学版), 2018, 52(12): 2372-2381.
[4] 何杰光, 彭志平, 崔得龙, 李启锐. 局部维度改进的教与学优化算法[J]. 浙江大学学报(工学版), 2018, 52(11): 2159-2170.
[5] 李志, 单洪, 马涛, 黄郡. 基于反向标签传播的移动终端用户群体发现[J]. 浙江大学学报(工学版), 2018, 52(11): 2171-2179.
[6] 王硕朋, 杨鹏, 孙昊. 听觉定位数据库构建过程优化[J]. 浙江大学学报(工学版), 2018, 52(10): 1973-1979.
[7] 魏小峰, 程承旗, 陈波, 王海岩. 基于独立边数的链码方法[J]. 浙江大学学报(工学版), 2018, 52(9): 1686-1693.
[8] 陈荣华, 王鹰汉, 卜佳俊, 于智, 高斐. 基于KNN算法与局部回归的网站无障碍采样评估[J]. 浙江大学学报(工学版), 2018, 52(9): 1702-1708.
[9] 张承志, 冯华君, 徐之海, 李奇, 陈跃庭. 图像噪声方差分段估计法[J]. 浙江大学学报(工学版), 2018, 52(9): 1804-1810.
[10] 刘洲洲, 李士宁, 李彬, 王皓, 张倩昀, 郑然. 基于弹性碰撞优化算法的传感云资源调度[J]. 浙江大学学报(工学版), 2018, 52(8): 1431-1443.
[11] 王勇超, 祝凯林, 吴奇轩, 鲁东明. 基于局部渲染的高精度模型自适应展示技术[J]. 浙江大学学报(工学版), 2018, 52(8): 1461-1466.
[12] 孙念, 李玉强, 刘爱华, 刘春, 黎威威. 基于松散条件下协同学习的中文微博情感分析[J]. 浙江大学学报(工学版), 2018, 52(8): 1452-1460.
[13] 郑守国, 崔雁民, 王青, 杨飞, 程亮. 飞机装配现场数据采集平台设计[J]. 浙江大学学报(工学版), 2018, 52(8): 1526-1534.
[14] 毕晓君, 王朝. 基于超平面投影的高维多目标进化算法[J]. 浙江大学学报(工学版), 2018, 52(7): 1284-1293.
[15] 张廷蓉, 滕奇志, 李征骥, 卿粼波, 何小海. 岩心三维CT图像超分辨率重建[J]. 浙江大学学报(工学版), 2018, 52(7): 1294-1301.