吉林大学学报(工学版) ›› 2017, Vol. 47 ›› Issue (2): 384-391.doi: 10.13229/j.cnki.jdxbgxb201702006

• • 上一篇    下一篇

基于卷积神经网络的道路车辆检测方法

李琳辉1, 2, 伦智梅1, 2, 连静1, 2, 袁鲁山1, 2, 周雅夫1, 2, 麻笑艺1, 2   

  1. 1.大连理工大学 工业装备结构分析国家重点实验室,辽宁 大连 116024;
    2.大连理工大学 汽车工程学院,辽宁 大连 116024
  • 收稿日期:2015-11-17 出版日期:2017-03-20 发布日期:2017-03-20
  • 通讯作者: 连静(1980-),女,副教授,博士.研究方向:汽车电子与控制.E-mail:lianjing80@126.com
  • 作者简介:李琳辉(1981-),男,副教授,博士.研究方向:智能车辆.E-mail:lilinhui@dlut.edu.cn
  • 基金资助:
    国家自然科学基金项目(61473057,61203171); 中央高校基本科研业务费专项项目(DUT15LK13).

Convolution neural network-based vehicle detection method

LI Lin-hui1, 2, LUN Zhi-mei1, 2, LIAN Jing1, 2, YUAN Lu-shan1, 2, ZHOU Ya-fu1, 2, MA Xiao-yi1, 2   

  1. 1.State Key Laboratory of Structural Analysis for Industrial Equipment, Dalian University of Technology, Dalian 116024, China;
    2.School of Automotive Engineering, Dalian University of Technology, Dalian 116024, China
  • Received:2015-11-17 Online:2017-03-20 Published:2017-03-20

摘要: 提出了一种基于卷积神经网络的前方车辆检测方法。首先,根据车底阴影特征,运用基于边缘增强的路面检测算法以及车底阴影自适应分割算法来分割并形成车底候选区域,以解决路面灰度分布不均及光照条件变化问题;其次,运用针对道路交通环境的卷积神经网络结构,建立图像样本库进行网络训练;在此基础上,采用基于卷积神经网络识别的方法以验证并剔除被误检测为车底阴影的候选区域,进而确定真正的车辆目标;最后,修改网络为三分类识别,以验证本文方法的强扩展性的优势。实验结果表明:本文提出的车辆检测方法能够很好地区分车底阴影和非车底阴影干扰,有效地提高车辆检测的准确率和可靠性,降低误检率。

关键词: 车辆工程, 车辆检测, 单目视觉, 卷积神经网络, 阴影分割

Abstract: A Convolution Neural Network (CNN) based vehicle detection method was proposed. First, an edge enhancement-based road detection and an adaptive shadow segmentation approaches are put forward to resolve the grayscale variation on the road and reduce the influence of the lighting variation. Then, the CNN structure applied to the road traffic environment is determined to train the established sample sets. On the basis of this, the vehicle region, which is wrongly detected as the vehicle shadow, is recognized by CNN and removed from the preliminary detection results, thus, the final vehicle shadow is obtained. Finally, CNN is modified to three classification to verify the advantages of a strong expandability of this method. The experimental results show that this method is effective in various conditions and meets the accuracy requirement, decreasing the false positive rate.

Key words: vehicle engineering, vehicle detection, monocular vision, convolutional neural network (CNN), shadow segmentation

中图分类号: 

  • U495
[1] Teoh S S,Bräunl T.Symmetry-based monocular vehicle detection system[J].Machine Vision and Applications,2012,23(5):831-842.
[2] Adhikari S P,Cho H,Yoo H J,et al.On-road succeeding vehicle detection using characteristic visual features[J].Transactions of the Korean Institute of Electrical Engineers,2010,59(3):636-644.
[3] Lalimi M A,Ghofrani S,Mclernon D.A vehicle license plate detection method using region and edge based methods[J].Computers & Electrical Engineering,2013,39(3):834-845.
[4] Kuo Y C,Pai N S,Li Y F.Vision-based vehicle detection for a driver assistance system[J].Computers & Mathematics with Applications,2011,61(8):2096-2100.
[5] Jia Bao-zhi,Liu Rui, Zhu Ming. Real-time obstacle detection with motion features using monocular vision[J].The Visual Computer,2015,31(3):281-293.
[6] 李云翀,何克忠,贾培发.基于阴影特征和Adaboost的前向车辆检测系统[J].清华大学学报:自然科学版,2007,47(10):1713-1716.
Li Yun-chong,He Ke-zhong,Jia Pei-fa.Forward vehicle detection based on shadow features and adaboost[J].J Tsinghua Univ (Sci & Tech),2007,47(10):1713-1716.
[7] Choi M K,Park J ,Lee S C.Event classification for vehicle navigation system by regional optical flow analysis[J].Machine Vision and Applications,2014,25(3):547-559.
[8] Álvarez S,Sotelo M Á,Ocaña M, et al.Perception advances in outdoor vehicle detection for automatic cruise control[J]. Robotica,2010, 28(5):765-779.
[9] 顾兆伦,邵雨辰,谢扬振,等.多环境下的实时前车检测与车距测量[J]. 信号处理,2015,31(9):1188-1194.
Gu Zhao-lun,Shao Yu-chen,Xie Yang-zhen, et al.Real-time vehicle detection and distance measurement in multi-environment[J].Journal of Signal Processing,2015,31(9):1188-1194.
[10] 金立生,王岩,刘景华,等.基于Adaboost算法的日间前方车辆检测[J].吉林大学学报:工学版,2014,44(6):1604-1608.
Jin Li-sheng,Wang Yan,Liu Jing-hua,et al. Front vehicle detection based on Adaboost algorithm in daytime[J].Journal of Jilin University (Engineering and Technology Edition),2014,44(6):1604-1608.
[11] Khammari A,Nashashibi F, Abramson Y,et al.Vehicle detection combining gradient analysis and AdaBoost classification[C]∥Proceedings of the 8th International IEEE Conference on Intelligent Transportation Systems,Vienna,Austria,2005:1084-1089.
[12] Sivaraman S,Trivedi M M. Active learning for on-road vehicle detection: a comparative study[J]. Machine Vision and Applications,2014,25(3):599-611.
[13] LeCun Y,Bengio Y,Hinton G.Deep learning[J]. Nature,2015,521(7553):436-444.
[14] Dan C, Meier U, Masci J, et al. Multi-column deep neural network for traffic sign classification[J]. Neural Networks,2012,32(1):333-338.
[15] Sermanet P, LeCun Y.Traffic sign recognition with multi-scale convolutional networks[C]∥Proc International Joint Conference on Neural Networks,San Jose,2011:2809-2813.
[16] Taigman Y,Yang M, Ranzato M,et al.Deepface: closing the gap to human-level performance in face verification[C]∥IEEE Conference on Computer Vision & Pattern Recognition, Columbus,2014:1701-1708.
[17] Ding C,Tao D.Robust face recognition via multimodal deep face representation[J].IEEE Transactions on Multimedia,2015,17(11):2049-2058.
[18] Farabet C,Couprie C,Najman L,et al.Learning hierarchical features for scene labeling[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence,2013,35(8):1915-1929.
[19] Xu K,Ba J,Kiros R, et al. Show, attend and tell: neural image caption generation with visual attention[C]∥Proc International Conference on Learning Representations,Lille, France,2015: 2048-2057.
[20] Vinyals O, Toshev A, Bengio S, et al. Show and tell: a neural image caption generator[C]∥IEEE Conference on Computer Vision and Pattern Recognition(CVPR), Boston,2015:3156-3164.
[21] Tanaka M, Morie T. Shadow detection and elimination using a light-source color vector and its application to in-vehicle camera images[J]. International Journal of Innovative Computing, Information and Control, 2015, 11(3): 865-879.
[22] Lee D U,Cheung R C C,Villasenor J D. A flexible architecture for precise Gamma correction[J]. IEEE Transactions on Very Large Scale Integration System,2007, 15(4): 474-478.
[1] 常成,宋传学,张雅歌,邵玉龙,周放. 双馈电机驱动电动汽车变频器容量最小化[J]. 吉林大学学报(工学版), 2018, 48(6): 1629-1635.
[2] 席利贺,张欣,孙传扬,王泽兴,姜涛. 增程式电动汽车自适应能量管理策略[J]. 吉林大学学报(工学版), 2018, 48(6): 1636-1644.
[3] 何仁,杨柳,胡东海. 冷藏运输车太阳能辅助供电制冷系统设计及分析[J]. 吉林大学学报(工学版), 2018, 48(6): 1645-1652.
[4] 那景新,慕文龙,范以撒,谭伟,杨佳宙. 车身钢-铝粘接接头湿热老化性能[J]. 吉林大学学报(工学版), 2018, 48(6): 1653-1660.
[5] 刘玉梅,刘丽,曹晓宁,熊明烨,庄娇娇. 转向架动态模拟试验台避撞模型的构建[J]. 吉林大学学报(工学版), 2018, 48(6): 1661-1668.
[6] 徐岩,孙美双. 基于卷积神经网络的水下图像增强方法[J]. 吉林大学学报(工学版), 2018, 48(6): 1895-1903.
[7] 赵伟强, 高恪, 王文彬. 基于电液耦合转向系统的商用车防失稳控制[J]. 吉林大学学报(工学版), 2018, 48(5): 1305-1312.
[8] 宋大凤, 吴西涛, 曾小华, 杨南南, 李文远. 基于理论油耗模型的轻混重卡全生命周期成本分析[J]. 吉林大学学报(工学版), 2018, 48(5): 1313-1323.
[9] 朱剑峰, 张君媛, 陈潇凯, 洪光辉, 宋正超, 曹杰. 基于座椅拉拽安全性能的车身结构改进设计[J]. 吉林大学学报(工学版), 2018, 48(5): 1324-1330.
[10] 那景新, 浦磊鑫, 范以撒, 沈传亮. 湿热环境对Sikaflex-265铝合金粘接接头失效强度的影响[J]. 吉林大学学报(工学版), 2018, 48(5): 1331-1338.
[11] 王炎, 高青, 王国华, 张天时, 苑盟. 混流集成式电池组热管理温均特性增效仿真[J]. 吉林大学学报(工学版), 2018, 48(5): 1339-1348.
[12] 金立生, 谢宪毅, 高琳琳, 郭柏苍. 基于二次规划的分布式电动汽车稳定性控制[J]. 吉林大学学报(工学版), 2018, 48(5): 1349-1359.
[13] 隗海林, 包翠竹, 李洪雪, 李明达. 基于最小二乘支持向量机的怠速时间预测[J]. 吉林大学学报(工学版), 2018, 48(5): 1360-1365.
[14] 王德军, 魏薇郦, 鲍亚新. 考虑侧风干扰的电子稳定控制系统执行器故障诊断[J]. 吉林大学学报(工学版), 2018, 48(5): 1548-1555.
[15] 胡满江, 罗禹贡, 陈龙, 李克强. 基于纵向频响特性的整车质量估计[J]. 吉林大学学报(工学版), 2018, 48(4): 977-983.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!