计算机科学 ›› 2019, Vol. 46 ›› Issue (11A): 309-314.

• 模式识别与图像处理 • 上一篇    下一篇

基于ST-CNN的交通标志实时检测识别算法

曲佳博, 秦勃   

  1. (中国海洋大学计算机科学与技术系 山东 青岛266100)
  • 出版日期:2019-11-10 发布日期:2019-11-20
  • 通讯作者: 秦勃(1964-),男,博士,教授,主要研究方向为计算机图形图像处理、并行计算、云计算。
  • 作者简介:曲佳博(1993-),男,硕士生,主要研究方向为计算机图形图像处理、并行计算,E-mail:13210000374@163.com。

Real-time Detection and Recognition Algorithm of Traffic Signs Based on ST-CNN

QU Jia-bo, QIN Bo   

  1. (Department of Computer Science & Technology,Ocean University of China,Qingdao,Shandong 266100,China)
  • Online:2019-11-10 Published:2019-11-20

摘要: 深度学习是基于图像的交通标志检测和识别处理的研究热点,已取得了显著的效果。针对基于车载视频的交通标志检测和识别处理问题,文中根据图像序列的帧间时空连续关系构建了时空关系模型(Spatiotemporal Model,STM),并与多尺度卷积神经网络(Convolutional Neural Networks,CNN)相结合,提出了一种基于时空卷积神经网络(Spatiotemporal-CNN,ST-CNN)的交通标志实时检测识别算法。实验结果表明,该算法可对视频图像序列中的同一交通标志实现检测、筛选、追踪和识别处理,在保证高准确率的同时,可有效减少CNN的数据输入,降低系统资源占用量,提高计算效率,满足了视频中交通标志检测识别的实时性要求。算法平均每帧耗时26.82ms,且识别准确率达到96.94%。

关键词: 多尺度卷积神经网络, 交通标志, 时空关系模型, 实时性

Abstract: At present,deep learning is a research hotspot based on image traffic sign detection and recognition proces-sing,and has achieved remarkable results.Aiming at the problem of traffic sign detection and recognition based on car-video,this paper proposed a real-time detection and recognition algorithm for traffic signs based on Spatiotemporal-CNN (ST-CNN).It constructs a Spatiotemporal model (STM) based on the spatiotemporal relationship between frames of image sequences,and combines the STM with Convolutional Neural Network (CNN).The experimental results show that the algorithm can detect,screen,track and identify the same traffic sign in the video image sequence.It can effectively reduce CNN data input and system resource consumption,and improve computational efficiency,while ensuring high accuracy.It satisfies the real-time requirements of traffic sign detection and recognition in video.The algorithm takes an average of 26.82 milliseconds per frame and the recognition accuracy reaches 96.94%.

Key words: Multi-scale convolutional neural network, Real-time, Spatiotemporal model, Traffic sign

中图分类号: 

  • TP391
[1]LUETTEL T,HIMMELSBACH M,WUENSCHE H J.Autonomous Ground Vehicles—Concepts and a Path to the Future [J].Proceedings of the IEEE,2012,100(5):1831-1839.
[2]赵娜,袁家斌,徐晗.智能交通系统综述[J].计算机科学,2014,41(11):7-11.
[3]LIU W,ANGUELOV D,ERHAN D,et al.SSD:Single Shot MultiBox Detector [M]∥Computer Vision-ECCV 2016.Chain:Springer,2016.
[4]REDMON J,DIVVALA S,GIRSHICK R,et al.You Only Look Once:Unified,Real-Time Object Detection [J].arXiv:1506.02640,2015.
[5]GAO X W,PODLADCHIKOVA L N,SHAPOSHNIKOV D G,et al.Recognition of traffic signs based on their colour and shape features extracted using human vision models[J].Journal of Visual Communication & Image Representation,2006,17(4):675-685.
[6]KUO W J,LIN C C.Two-Stage Road Sign Detection and Recognition[C]∥2007 IEEE International Conference on Multimedia and Expo.IEEE,2007.
[7]JIANG Y,ZHOU S,JIANG Y,et al.Traffic sign recognitionusing Ridge Regression and OTSU method[C]∥Intelligent Vehicles Symposium.IEEE,2011.
[8]UIJLINGS J R R,VAN DE SANDE K E A.Selective Search for Object Recognition [J].International Journal of Computer Vision,2013,104(2):154-171.
[9]REN S,HE K,GIRSHICK R,et al.Faster R-CNN:Towards Real-Time Object Detection with Region Proposal Networks [J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2015,39(16):1137-1149.
[10]HE K,GKIOXARI G,DOLLÁR P,et al.Mask R-CNN [J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2017,PP(99):1-1.
[11]STALLKAMP J,SCHLIPSING M,SALMEN J,et al.Man vs.computer:Benchmarking machine learning algorithms for traffic sign recognition [J].Neural newtorks:the official Journal of the International Neural Network Society,2012,32(2):323-332.
[12]WU Y,LIU Y,LI J,et al.Traffic sign detection based on convolutional neural networks[C]∥International Joint Conference on Neural Networks.IEEE,2014.
[13]GREENHALGH J,MIRMEHDI M.Real-Time Detection andRecognition of Road Traffic Signs [J].IEEE Transactions on Intelligent Transportation Systems,2012,13(4):1498-1506.
[14]IGEL C.Detection of traffic signs in real-world images:the German traffic sign detection benchmark[C]∥International Joint Conference on Neural Networks.IEEE,2013.
[15]WANG G,REN G,WU Z,et al.A robust,coarse-to-fine traffic sign detection method[C]∥International Joint Conference on Neural Networks.IEEE,2013.
[16]STALLKAMP J,SCHLIPSING M,et al.The German Traffic Sign Recognition Benchmark:A multi-class classification competition[C]∥International Joint Conference on Neural Networks.IEEE,2011.
[17]LU X,WANG Y,ZHOU X,et al.Traffic Sign Recognition via Multi-Modal Tree-Structure Embedded Multi-Task Learning [J].IEEE Transactions on Intelligent Transportation Systems,2017,18(4):960-972.
[18]LUO H,YANG Y,TONG B,et al.Traffic Sign RecognitionUsing a Multi-Task Convolutional Neural Network [J].IEEE Transactions on Intelligent Transportation Systems,2018,PP (99):1-12.
[19]SERMANET P,LECUN Y.Traffic Sign Recognition withMulti-Scale Convolutional Networks[C]∥The 2011 International Joint Conference on Neural Networks (IJCNN).2011.
[1] 胡聪, 何晓晖, 邵发明, 张艳武, 卢冠林, 王金康.
基于极大极稳定区域及SVM的交通标志检测
Traffic Sign Detection Based on MSERs and SVM
计算机科学, 2022, 49(6A): 325-330. https://doi.org/10.11896/jsjkx.210300117
[2] 吴培培, 吴兆贤, 唐文兵.
基于吸收态马尔可夫链的智能无人车系统实时性能分析
Real-time Performance Analysis of Intelligent Unmanned Vehicle System Based on Absorbing Markov Chain
计算机科学, 2021, 48(11A): 147-153. https://doi.org/10.11896/jsjkx.210300050
[3] 张元鸣, 李梦妮, 黄浪游, 陆佳炜, 肖刚.
基于增量日志的数据组合视图定位更新方法
Data Composition View Positioning Update Approach with Incremental Logs
计算机科学, 2020, 47(6): 85-91. https://doi.org/10.11896/jsjkx.190500085
[4] 庞宇, 刘平, 雷印杰.
基于移动端的“非受控”物体识别算法的实现
Realization of “Uncontrolled” Object Recognition Algorithm Based on Mobile Terminal
计算机科学, 2019, 46(6A): 153-157.
[5] 李宗鑫, 秦勃, 王梦倩.
基于时空关系模型的交通信号灯的实时检测与识别
Real-time Detection and Recognition of Traffic Light Based on Time-Space Model
计算机科学, 2018, 45(6): 314-319. https://doi.org/10.11896/j.issn.1002-137X.2018.06.055
[6] 黄中平,白光伟,沈航,承骁,华志翔.
基于MapReduce模型的推测执行优化算法
Speculative Execution Optimization Algorithm with MapReduce
计算机科学, 2017, 44(4): 193-196. https://doi.org/10.11896/j.issn.1002-137X.2017.04.042
[7] 孙磊,杨海燕,吴际.
基于IMA平台的嵌入式软件设计模型仿真及实时性分析方法
Simulation and Real-time Analysis for Embedded Software Design Model with Consideration of Integrated Modular Avionics Platform
计算机科学, 2015, 42(12): 95-97.
[8] 郑远力,胡志坤.
基于滑动扫描框的高速物体的图像实时跟踪算法
Real-time Tracking Algorithm for Fast Target Based on Dynamical Scanning Boxes
计算机科学, 2015, 42(10): 287-291.
[9] 王棚飞,刘宏哲,袁家政,陈丽.
一种基于图像匹配的地面交通标志实时识别方法
Method of Real Time Recognition of Ground Traffic Signs Based on Image Matching
计算机科学, 2014, 41(6): 317-325. https://doi.org/10.11896/j.issn.1002-137X.2014.06.064
[10] 柏骏,夏靖波,吴吉祥,任高明,赵小欢.
实时网络流量分类研究综述
Survey on Real-time Traffic Classification
计算机科学, 2013, 40(9): 8-15.
[11] 余云霞,綦志勇.
基于混沌反馈控制理论的资源选择算法研究
Research on Resource Selection Algorithm Based on Feedback Control of Chaos Theory
计算机科学, 2012, 39(Z6): 452-456.
[12] 李 允,桂盛霖,陈 更,罗 蕾.
嵌入式实时软件模型开发环境研究
Model Development Environment Research of Embedded Real-time Software
计算机科学, 2012, 39(Z11): 226-229.
[13] 唐国明,周广新,谢羿,汤大权,唐九阳.
一种基于双层栅格划分的无线传感器网络目标定位方法
Target Localization Based on Double-level Grid Division in Wireless Sensor Networks
计算机科学, 2012, 39(6): 25-29.
[14] 周伟,安虹,刘谷,李小强,吴石磊.
一种输入感知的雷达回波快速聚类实现
Input-aware Runtime Scheduling Support for Fast Clustering of Radar Reflectivity Data on GPUs
计算机科学, 2012, 39(12): 295-299.
[15] 范荣全,肖红,李琪林.
智能变电站GOOSE通信网实时性分析
Real-time Performance Analysis of GOOSE Communication Network in Smart Substation
计算机科学, 2011, 38(Z10): 444-446.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!