基于移动小车的行人监控系统

doi:10.11959/j.issn.1000-0801.2017042

摘要/Abstract

摘要：

行人监控属于监控系统中比较重要的一个方面。传统的行人监控手段存在监控区域受限的问题，即监控设备一旦安装后，只能监控固定的区域，目前仍然需要人工判别监控画面中是否有行人。针对这个问题，设计了基于计算机视觉的移动行人监控系统。通过操控搭载了监控摄像头的移动小车，实现监控区域的自由切换。另外，配合成熟的深度卷积神经网络算法和相关滤波器，实现了监控画面中行人的自主识别、定位。最后对该系统进行了实地测试，验证了行人监控系统的可行性。

关键词: 计算机视觉, 行人监控, 移动小车, 深度卷积神经网络, 相关滤波器

Abstract:

Pedestrian surveillance is one of the most important aspects in the surveillance system.Traditional surveillance equipment can only cover the limit area.Namely,once the surveillance equipment is implemented,it can only monitor a specific region.Besides,the surveillance system can't detect if there is a person in the surveillance picture or not.The professional staffs are needed to determine whether there is a person in the surveillance picture.To solve this problem,a computer vision based on mobile pedestrian surveillance system was designed.By implementing a surveillance camera on a mobile small vehicle and remotely control it,the system could switch the monitoring area.Besides,the system could classify and locate the pedestrian in the picture with the deep convolutional neural network and correlation filters.Finally,the system was tested on the spot,which verified the feasibility of the system.

Key words: computer vision, pedestrian surveillance, mobile vehicle, deep convolutional neural network, correlation filter

中图分类号:

TP277

邢惠钧,昌硕. 基于移动小车的行人监控系统[J]. 电信科学, 2017, 33(2): 120-127.

Huijun XING,Shuo CHANG. Pedestrian surveillance system based on mobile vehicle[J]. Telecommunications Science, 2017, 33(2): 120-127.

图/表 10

图1

图2

图3

图5

图4

图6

图7

表1

图8

图9

参考文献 24

[1]	DALAL N , TRIGGS B . Histograms of oriented gradients for human detection[C]// Computer Vision and Pattern Recognition, June 20-26, 2005, San Diego,CA,USA. New Jersey: IEEE Press, 2005:886-893.
[2]	SUYKENS J A , VANDEWALLE J . Least squares support vector machine classifiers[J]. Neural Processing Letters, 1999,9(3):293-300.
[3]	ZHANG L , LIN L , LIANG X , et al. Is faster R-CNN doing well for pedestrian detection[C]// European Conference on Computer Vision, October 8-16, 2016, Amsterdam,Netherlands. Berlin: Springe, 2016:443-457.
[4]	SOLANKI D K M S . Pedestrian detection using R-CNN[J]. Group, 2016,12228(2):12419.
[5]	OLIVEIRA L , NUNES U , PEIXOTO P . On exploration of classifier ensemble synergism in pedestrian detection[J]. IEEE Transactions on Intelligent Transportation Systems, 2010,11(1):16-27.
[6]	QUYANG W , WANG X . Joint deep learning for pedestrian detection[C]// International Conference on Computer Vision, Dec 1-8, 2013, Sydney,Australia. New Jersey: IEEE Press, 2013:2056-2063.
[7]	LUO P , TIAN Y , WANG X , et al. Switchable deep network for pedestrian detection[C]// Computer Vision and Pattern Recognition, Jun 23-28, 2014, OH,USA. Berlin: IEEE Press, 2014:899-906.
[8]	SERMANET P , KAVUKCUOGLU K , CHINTALA S , et al. Pedestrian detection with unsupervised multi-stage feature learning[C]// Computer Vision and Pattern Recognition, Jun 23-28, 2013, ORUSA. New Jersey: IEEE Press, 2013:3626-3633.
[9]	GIRSHICK R . Fast R-CNN[C]// International Conference on Computer Vision, Jun 7-12, 2015, MA,USA. New Jersey: IEEE Press, 2015:1440-1448.
[10]	REN S , HE K , GIRSHICK R , et al. Faster R-CNN:towards real-time object detection with region proposal networks[C]// Advances in Neural Information Processing Systems, December 7-12, 2015, Montreal,Canada. New York: Curran Associates, 2015:91-99.
[11]	DOLLáR P , WOJEK C , SCHIELE B , et al. Pedestrian detection:a benchmark[C]// Computer Vision and Pattern Recognition, June 20-25, 2009, FL,USA. New Jersey: IEEE Press, 2009:304-311.
[12]	HENRIQUES J.F , CASEIRO R , MARTINS P , et al. High-speed tracking with kernelized correlation filters[J]. Pattern Analysis and Machine Intelligence, 2015,37(3):583-596.
[13]	SIMONYAN K , ZISSERMAN A . Very deep convolutional networks for large-scale image recognition[J]. Computing Research Repository, 2014,abs:1409-1556.
[14]	NAIR V , HINTON G.E . Rectified linear units improve restricted Boltzmann machines[C]// International Conference on Machine Learning, Dec 12-14, 2010, Washington,DC,USA. New Jersey: IEEE Press, 2010:807-814.
[15]	SUTSKEVER I , MARTENS J , DAHL G E , et al. On the importance of initialization and momentum in deep learning[J]. ICML, 2013,(28):1139-1147.
[16]	PITOR D , CHRISTIAN W , BERNT S , et al. Pedestrian detection:an evaluation of the state of the art[J]. PAMI, 2012,34(4):743-761.
[17]	RUSSAKOVSKY O , DENG J , SU H , et al. Imagenet large scale visual recognition challenge[J]. International Journal of Computer Vision, 2015,115(3):211-252.
[18]	BOLME , DAVID S , et al. Visual object tracking using adaptive correlation filters[C]// Computer Vision and Pattern Recognition, Dec 12-14, 2010, Washington,DC,USA. New Jersey: IEEE Press, 2010:2544-2550.
[19]	HENRIQUES , CASEIRO , et al. Exploiting the circulant structure of tracking-by-detection with kernels[C]// European Conference on Computer Vision, Oct 7, 2012, Firenze,Italy. Berlin: Springer, 2012:702-715.
[20]	VIOLA P , JONES M J . Robust real-time face detection[J]. International Journal of Computer Vision, 2004,57(2):137-154.
[21]	WALK S , MAJER N , SCHINDLER K , et al. New features and insights for pedestrian detection[C]// 2010 IEEE Conference on Computer Vis ion and Pattern Recognition, Dec 12-14, 2010, Washington,DC,USA. New Jersey: IEEE Press, 2010:1030-1037.
[22]	NAM W , DOLLAR P , HAN J H . Local decorrelation for improved pedestrian detection[C]// Advances in Neural Information Processing Systems, Dec 8-13, 2014, Montréal,CANADA. New York: Curran Associates, 2014:424-432.
[23]	HOSANG J , OMRAN M , BENENSON R , et al. Taking a deeper look at pedestrians[C]// 2015 Computer Vision and Pattern Recognition, Jun 8-12, 2015, Boston,MA,USA. New Jersey: IEEE Press, 2015:4073-4082.
[24]	LI J , LIANG X , SHEN S M , et al. Scale-aware fast R-CNN for pedestrian detection[J]. arXiv preprint arXiv:1510.08160, 2015.

网络名称	测试时间/s	%漏检率
本文提出模型	0.26	23%
SAF R-CNN	0.37	10%
R-CNN	5.31	13%