Real-time people counting using blob descriptor

Research output: Contribution to journalConference article

18 Citations (Scopus)

Abstract

We propose a system for counting the number of pedestrians in real-time. This system estimates "how many pedestrians are and where they are in video sequences" by the following procedures. First, candidate regions are segmented into blobs according to background subtraction. Second, a set of features are extracted from each blob and a neural network estimates the number of pedestrians corresponding to each set of features. To realize real-time processing, we used only simple and valid features, and the adaptive background modeling using Parzen density estimation, which realizes fast and accurate object detection in input images. We also validate the effectiveness of the proposed system by several experiments.

Original languageEnglish
Pages (from-to)143-152
Number of pages10
JournalProcedia - Social and Behavioral Sciences
Volume2
Issue number1
DOIs
Publication statusPublished - Aug 2 2010
Event1st International Conference on Security Camera Network, Privacy Protection and Community Safety 2009, SPC2009 - Kiryu, Japan
Duration: Oct 28 2009Oct 30 2009

Fingerprint

pedestrian
neural network
candidacy
video
experiment
time
Pedestrians

All Science Journal Classification (ASJC) codes

  • Social Sciences(all)
  • Psychology(all)

Cite this

Real-time people counting using blob descriptor. / Yoshinaga, Satoshi; Shimada, Atsushi; Taniguchi, Rin-Ichiro.

In: Procedia - Social and Behavioral Sciences, Vol. 2, No. 1, 02.08.2010, p. 143-152.

Research output: Contribution to journalConference article

@article{74650272b1e1471e83cade8eb913edab,
title = "Real-time people counting using blob descriptor",
abstract = "We propose a system for counting the number of pedestrians in real-time. This system estimates {"}how many pedestrians are and where they are in video sequences{"} by the following procedures. First, candidate regions are segmented into blobs according to background subtraction. Second, a set of features are extracted from each blob and a neural network estimates the number of pedestrians corresponding to each set of features. To realize real-time processing, we used only simple and valid features, and the adaptive background modeling using Parzen density estimation, which realizes fast and accurate object detection in input images. We also validate the effectiveness of the proposed system by several experiments.",
author = "Satoshi Yoshinaga and Atsushi Shimada and Rin-Ichiro Taniguchi",
year = "2010",
month = "8",
day = "2",
doi = "10.1016/j.sbspro.2010.01.028",
language = "English",
volume = "2",
pages = "143--152",
journal = "Procedia - Social and Behavioral Sciences",
issn = "1877-0428",
publisher = "Elsevier BV",
number = "1",

}

TY - JOUR

T1 - Real-time people counting using blob descriptor

AU - Yoshinaga, Satoshi

AU - Shimada, Atsushi

AU - Taniguchi, Rin-Ichiro

PY - 2010/8/2

Y1 - 2010/8/2

N2 - We propose a system for counting the number of pedestrians in real-time. This system estimates "how many pedestrians are and where they are in video sequences" by the following procedures. First, candidate regions are segmented into blobs according to background subtraction. Second, a set of features are extracted from each blob and a neural network estimates the number of pedestrians corresponding to each set of features. To realize real-time processing, we used only simple and valid features, and the adaptive background modeling using Parzen density estimation, which realizes fast and accurate object detection in input images. We also validate the effectiveness of the proposed system by several experiments.

AB - We propose a system for counting the number of pedestrians in real-time. This system estimates "how many pedestrians are and where they are in video sequences" by the following procedures. First, candidate regions are segmented into blobs according to background subtraction. Second, a set of features are extracted from each blob and a neural network estimates the number of pedestrians corresponding to each set of features. To realize real-time processing, we used only simple and valid features, and the adaptive background modeling using Parzen density estimation, which realizes fast and accurate object detection in input images. We also validate the effectiveness of the proposed system by several experiments.

UR - http://www.scopus.com/inward/record.url?scp=77954960555&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77954960555&partnerID=8YFLogxK

U2 - 10.1016/j.sbspro.2010.01.028

DO - 10.1016/j.sbspro.2010.01.028

M3 - Conference article

AN - SCOPUS:77954960555

VL - 2

SP - 143

EP - 152

JO - Procedia - Social and Behavioral Sciences

JF - Procedia - Social and Behavioral Sciences

SN - 1877-0428

IS - 1

ER -