适用于密集人群目标检测的多尺度检测方法

系统仿真学报 ›› 2016, Vol. 28 ›› Issue (10): 2503-2509.

适用于密集人群目标检测的多尺度检测方法

周建新¹, 吴建军², 薛均强², 林帅¹, 党岗¹, 程志全^3,*

1.国防科学技术大学计算机学院,湖南长沙 410073;
2.中国人民解放军海南武警总队,海南海口 570203;
3.湖南化身科技有限公司,湖南长沙 410205

收稿日期:2016-05-04 修回日期:2016-08-04 出版日期:2016-10-08 发布日期:2020-08-13
作者简介:周建新(1989-),男,甘肃平凉,硕士,助工,研究方向为多媒体与虚拟现实技术;吴建军(1970-),男,海南海口,硕士,研究方向为视频处理与通信;薛均强(1982-),男,广东深圳,硕士,助工,研究方向为计算机图形学。

Multi-scale Detection Method for Dense Crowd Target Detection

Zhou Jianxin¹, Wu Jianjun², Xue Junqiang², Lin Shuai¹, Dang Gang¹, Cheng Zhiquan^3,*

1. National University of Defense Technology, Changsha 410073, China;
2. Hainan Armed Police Force, Hainan 570203, China;
3. Hunan Avatar Science Company, Changsha 410205, China

Received:2016-05-04 Revised:2016-08-04 Online:2016-10-08 Published:2020-08-13

摘要/Abstract

摘要： 针对密集人群场景下的目标检测问题,提出了一种多尺度的目标检测方法。在粗尺度下,使用优化的DPM(Deformable Part Model)检测方法,将人体全身作为检测对象,检测整个场景中的稀疏目标;在细尺度下,将头部作为检测对象,使用重新训练的Faster R-CNN(Region-based Convolutional Neural Network)网络检测稠密人群中的目标。将两种尺度下检测结果通过非极大值抑制(NMS,Non-Maximum Suppression)方法结合在一起,这样两种方法既互相补充又能去除冗余检测结果。实验结果证明,相比于单独的DPM检测方法和Faster R-CNN检测方法,提出的多尺度检测方法在检测精度上有显著提升。

关键词: 密集人群检测, 多尺度检测, DPM, Faster R-CNN

Abstract: A multi-scale algorithm was proposed to detect the targets flexibly. In coarse scale, an optimized DPM (Deformable Part Model) method was used to filter out sparse objectives that was represented by whole body. Then the whole scenario was cut into multiple finer regions and the Faster R-CNN (Region-based Convolutional Neural Network) method was trained and utilized to detect dense objects that was indicated by head in fine scale. These two detection results were incorporated via NMS (Non - Maximum Suppression) method, in order to supplement with each other and remove redundancy. The effectiveness of the proposed method has been proved comparing detect accuracy with DPM and R-CNN individually in the final experiment.

Key words: dense crowd target detection, multi-scale detection, DPM, Faster R-CNN

中图分类号:

TP18

周建新, 吴建军, 薛均强, 林帅, 党岗, 程志全. 适用于密集人群目标检测的多尺度检测方法[J]. 系统仿真学报, 2016, 28(10): 2503-2509.

Zhou Jianxin, Wu Jianjun, Xue Junqiang, Lin Shuai, Dang Gang, Cheng Zhiquan. Multi-scale Detection Method for Dense Crowd Target Detection[J]. Journal of System Simulation, 2016, 28(10): 2503-2509.

参考文献 20

[1]	Dalal N, Triggs B.Histograms of oriented gradients for human detection[C]// Computer Vision and Pattern Recognition, 2005. CVPR 2005, IEEE Computer Society Conference on. USA: IEEE, 2005, 1: 886-893.
[2]	Felzenszwalb P F, Girshick R B, McAllester D, et al. Object detection with discriminatively trained part-based models[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence (S0162-8828), 2010, 32(9): 1627-1645.
[3]	Felzenszwalb P F, Girshick R B, McAllester D. Cascade object detection with deformable part models[C]//Computer vision and pattern recognition (CVPR), 2010 IEEE conference on. USA:IEEE, 2010: 2241-2248.
[4]	Wang X, Yang M, Zhu S, et al.Regionlets for generic object detection[C]// Computer Vision (ICCV), 2013 IEEE International Conference on. USA: IEEE, 2013: 17-24.
[5]	Lempitsky V, Zisserman A.Learning to count objects in images[C]//Advances in Neural Information Processing Systems. Canada:NIPS,2010: 1324-1332.
[6]	Ren X, Ramanan D.Histograms of sparse codes for object detection[C]// Computer Vision and Pattern Recognition (CVPR), 2013 IEEE Conference on. USA: IEEE, 2013: 3246-3253.
[7]	Haroon Idrees, Imran Saleemi, Cody Seibert, Mubarak Shah.Muti-Source Multi-Scale Counting in Extremely Dense Crowd Images[C]// The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2013. USA: IEEE, 2013: 2547-2554.
[8]	Bansal A, Venkatesh K S. People Counting in High Density Crowds from Still Images [J]. arXiv preprint arXiv (S2331-8422), 1507.08445, 2015. https://arxiv.org.
[9]	Krizhevsky A, Sutskever I, Hinton G E.Imagenet classification with deep convolutional neural networks[C]// Advances in Neural Information Processing Systems. Spain:NIPS , 2012: 1097-1105.
[10]	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. USA:IEEE,2014: 580-587.
[11]	Girshick R.Fast r-cnn[C]//Proceedings of the IEEE International Conference on Computer Vision. USA: IEEE, 2015: 1440-1448.
[12]	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.Canada:NIPS, 2015: 91-99.
[13]	Redmon J, Divvala S, Girshick R, et al. You only look once: Unified, real-time object detection [J]. arXiv preprint arXiv (S2331-8422), 1506.02640,2015. https://arxiv.org.
[14]	Wang C, Zhang H, Yang L, et al.Deep People Counting in Extremely Dense Crowds[C]//Proceedings of the 23rd ACM international conference on Multimedia. Australia,ACM, 2015: 1299-1302.
[15]	Zeiler M D, Fergus R.Visualizing and Understanding Convolutional Networks [M]// Computer Vision - ECCV 2014. Springer International Publishing, 2013:818-833.
[16]	Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition [J]. arXiv preprint arXiv (S2331-8422), 1409.1556, 2014. https://arxiv.org.
[17]	Rodriguez M, Laptev I, Sivic J, et al.Density-aware person detection and tracking in crowds[C]//2011 International Conference on Computer Vision. Spain:IEEE, 2011: 2423-2430.
[18]	Felzenszwalb P F, Huttenlocher D P.Pictorial structures for object recognition[J]. International Journal of Computer Vision (S0920-5691), 2005, 61(1): 55-79.
[19]	Fischler M A, Elschlager R A.The representation and matching of pictorial structures[J]. IEEE Transactions on computers (S0018-9340), 1973, 22(1): 67-92.
[20]	Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun.Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition[C]// IEEE Transactions on Pattern Analysis and Machine Intelligence. USA: IEEE, 2015: 1904-1916.