Pedestrian detection based on combining classifier in surveillance video

Yutang Wu; Xiaohua Wang; Haihong Wu

doi:10.1007/978-3-642-35211-9_63

Pedestrian detection based on combining classifier in surveillance video

Yutang Wu, Xiaohua Wang, Haihong Wu

Hangzhou Dianzi University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

In the field of visual surveillance, pedestrian detection can be used in many situations, which is concerned by many researchers. In order to solve the problem of partial occlusion, a pedestrian detection method based on head-shoulders features is proposed. Firstly, the location of pedestrian can be obtained roughly by using vertical and horizontal projection characteristics of the foreground pixels in the slid windows combined with AdaBoost classifier. Secondly, we can obtain the head-shoulders content correctly by using the histogram of edge orient gradient characteristics combined with SVM. The experimental results indicate that the proposed method is effective to solve partial occlusion problem.

Original language	English
Title of host publication	Network Computing and Information Security
Subtitle of host publication	Second International Conference, NCIS 2012 Shanghai, China, December 7-9, 2012 Proceedings
Editors	Fu LeeWang, Mo Li, Yuan Luo
Pages	492-500
Number of pages	9
DOIs	https://doi.org/10.1007/978-3-642-35211-9_63
Publication status	Published - 2012
Externally published	Yes
Event	2nd International Conference on Network Computing and Information Security, NCIS 2012 - Shanghai, China Duration: 7 Dec 2012 → 9 Dec 2012

Publication series

Name	Communications in Computer and Information Science
Volume	345
ISSN (Print)	1865-0929

Conference

Conference	2nd International Conference on Network Computing and Information Security, NCIS 2012
Country/Territory	China
City	Shanghai
Period	7/12/12 → 9/12/12

Keywords

AdaBoost
head-shoulders features
pedestrian detection
SVM

Access to Document

10.1007/978-3-642-35211-9_63

Cite this

Wu, Y., Wang, X., & Wu, H. (2012). Pedestrian detection based on combining classifier in surveillance video. In F. LeeWang, M. Li, & Y. Luo (Eds.), Network Computing and Information Security: Second International Conference, NCIS 2012 Shanghai, China, December 7-9, 2012 Proceedings (pp. 492-500). (Communications in Computer and Information Science; Vol. 345). https://doi.org/10.1007/978-3-642-35211-9_63

@inproceedings{cfaa3b589a884e0a94ee8a28e6b8c4c9,

title = "Pedestrian detection based on combining classifier in surveillance video",

abstract = "In the field of visual surveillance, pedestrian detection can be used in many situations, which is concerned by many researchers. In order to solve the problem of partial occlusion, a pedestrian detection method based on head-shoulders features is proposed. Firstly, the location of pedestrian can be obtained roughly by using vertical and horizontal projection characteristics of the foreground pixels in the slid windows combined with AdaBoost classifier. Secondly, we can obtain the head-shoulders content correctly by using the histogram of edge orient gradient characteristics combined with SVM. The experimental results indicate that the proposed method is effective to solve partial occlusion problem.",

keywords = "AdaBoost, head-shoulders features, pedestrian detection, SVM",

author = "Yutang Wu and Xiaohua Wang and Haihong Wu",

year = "2012",

doi = "10.1007/978-3-642-35211-9_63",

language = "English",

isbn = "9783642352102",

series = "Communications in Computer and Information Science",

pages = "492--500",

editor = "Fu LeeWang and { Li}, Mo and { Luo}, Yuan",

booktitle = "Network Computing and Information Security",

note = "2nd International Conference on Network Computing and Information Security, NCIS 2012 ; Conference date: 07-12-2012 Through 09-12-2012",

}

Wu, Y, Wang, X & Wu, H 2012, Pedestrian detection based on combining classifier in surveillance video. in F LeeWang, M Li & Y Luo (eds), Network Computing and Information Security: Second International Conference, NCIS 2012 Shanghai, China, December 7-9, 2012 Proceedings. Communications in Computer and Information Science, vol. 345, pp. 492-500, 2nd International Conference on Network Computing and Information Security, NCIS 2012, Shanghai, China, 7/12/12. https://doi.org/10.1007/978-3-642-35211-9_63

Pedestrian detection based on combining classifier in surveillance video. / Wu, Yutang; Wang, Xiaohua; Wu, Haihong.
Network Computing and Information Security: Second International Conference, NCIS 2012 Shanghai, China, December 7-9, 2012 Proceedings. ed. / Fu LeeWang; Mo Li; Yuan Luo. 2012. p. 492-500 (Communications in Computer and Information Science; Vol. 345).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Pedestrian detection based on combining classifier in surveillance video

AU - Wu, Yutang

AU - Wang, Xiaohua

AU - Wu, Haihong

PY - 2012

Y1 - 2012

N2 - In the field of visual surveillance, pedestrian detection can be used in many situations, which is concerned by many researchers. In order to solve the problem of partial occlusion, a pedestrian detection method based on head-shoulders features is proposed. Firstly, the location of pedestrian can be obtained roughly by using vertical and horizontal projection characteristics of the foreground pixels in the slid windows combined with AdaBoost classifier. Secondly, we can obtain the head-shoulders content correctly by using the histogram of edge orient gradient characteristics combined with SVM. The experimental results indicate that the proposed method is effective to solve partial occlusion problem.

AB - In the field of visual surveillance, pedestrian detection can be used in many situations, which is concerned by many researchers. In order to solve the problem of partial occlusion, a pedestrian detection method based on head-shoulders features is proposed. Firstly, the location of pedestrian can be obtained roughly by using vertical and horizontal projection characteristics of the foreground pixels in the slid windows combined with AdaBoost classifier. Secondly, we can obtain the head-shoulders content correctly by using the histogram of edge orient gradient characteristics combined with SVM. The experimental results indicate that the proposed method is effective to solve partial occlusion problem.

KW - AdaBoost

KW - head-shoulders features

KW - pedestrian detection

KW - SVM

UR - http://www.scopus.com/inward/record.url?scp=84882261566&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-35211-9_63

DO - 10.1007/978-3-642-35211-9_63

M3 - Conference contribution

AN - SCOPUS:84882261566

SN - 9783642352102

T3 - Communications in Computer and Information Science

SP - 492

EP - 500

BT - Network Computing and Information Security

A2 - LeeWang, Fu

A2 - Li, Mo

A2 - Luo, Yuan

T2 - 2nd International Conference on Network Computing and Information Security, NCIS 2012

Y2 - 7 December 2012 through 9 December 2012

ER -

Wu Y, Wang X, Wu H. Pedestrian detection based on combining classifier in surveillance video. In LeeWang F, Li M, Luo Y, editors, Network Computing and Information Security: Second International Conference, NCIS 2012 Shanghai, China, December 7-9, 2012 Proceedings. 2012. p. 492-500. (Communications in Computer and Information Science). doi: 10.1007/978-3-642-35211-9_63

Pedestrian detection based on combining classifier in surveillance video

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this