Partially occluded human detection by boosting SVM

Shaopeng Tang, Satoshi Goto

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

In this paper, a novel method to detect partially occluded humans in still images is proposed. An individual human is modeled as an assembly of natural body parts. Some part based SVM classifiers are trained first by using histogram of orientated gradient feature. Different from other boosting methods, region information is stored in each classifier. When detect human in crowed scene, according to the information of humans that have already been detected, the information of available regions could be obtained, when a new detection window is in process. In classifier sequence, the classifiers whose regions are available are selected for generating the final classifier. This method could achieve good performance on images and video sequences with several occlusions.

Original languageEnglish
Title of host publicationProceedings of 2009 5th International Colloquium on Signal Processing and Its Applications, CSPA 2009
Pages224-227
Number of pages4
DOIs
Publication statusPublished - 2009
Event2009 5th International Colloquium on Signal Processing and Its Applications, CSPA 2009 - Kuala Lumpur
Duration: 2009 Mar 62009 Mar 8

Other

Other2009 5th International Colloquium on Signal Processing and Its Applications, CSPA 2009
CityKuala Lumpur
Period09/3/609/3/8

Fingerprint

Classifiers

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Computer Science Applications
  • Computer Vision and Pattern Recognition
  • Signal Processing

Cite this

Tang, S., & Goto, S. (2009). Partially occluded human detection by boosting SVM. In Proceedings of 2009 5th International Colloquium on Signal Processing and Its Applications, CSPA 2009 (pp. 224-227). [5069221] https://doi.org/10.1109/CSPA.2009.5069221

Partially occluded human detection by boosting SVM. / Tang, Shaopeng; Goto, Satoshi.

Proceedings of 2009 5th International Colloquium on Signal Processing and Its Applications, CSPA 2009. 2009. p. 224-227 5069221.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Tang, S & Goto, S 2009, Partially occluded human detection by boosting SVM. in Proceedings of 2009 5th International Colloquium on Signal Processing and Its Applications, CSPA 2009., 5069221, pp. 224-227, 2009 5th International Colloquium on Signal Processing and Its Applications, CSPA 2009, Kuala Lumpur, 09/3/6. https://doi.org/10.1109/CSPA.2009.5069221
Tang S, Goto S. Partially occluded human detection by boosting SVM. In Proceedings of 2009 5th International Colloquium on Signal Processing and Its Applications, CSPA 2009. 2009. p. 224-227. 5069221 https://doi.org/10.1109/CSPA.2009.5069221
Tang, Shaopeng ; Goto, Satoshi. / Partially occluded human detection by boosting SVM. Proceedings of 2009 5th International Colloquium on Signal Processing and Its Applications, CSPA 2009. 2009. pp. 224-227
@inproceedings{0d251ee3412b4ffd972f1ac9ce0f14ea,
title = "Partially occluded human detection by boosting SVM",
abstract = "In this paper, a novel method to detect partially occluded humans in still images is proposed. An individual human is modeled as an assembly of natural body parts. Some part based SVM classifiers are trained first by using histogram of orientated gradient feature. Different from other boosting methods, region information is stored in each classifier. When detect human in crowed scene, according to the information of humans that have already been detected, the information of available regions could be obtained, when a new detection window is in process. In classifier sequence, the classifiers whose regions are available are selected for generating the final classifier. This method could achieve good performance on images and video sequences with several occlusions.",
author = "Shaopeng Tang and Satoshi Goto",
year = "2009",
doi = "10.1109/CSPA.2009.5069221",
language = "English",
isbn = "9781424441501",
pages = "224--227",
booktitle = "Proceedings of 2009 5th International Colloquium on Signal Processing and Its Applications, CSPA 2009",

}

TY - GEN

T1 - Partially occluded human detection by boosting SVM

AU - Tang, Shaopeng

AU - Goto, Satoshi

PY - 2009

Y1 - 2009

N2 - In this paper, a novel method to detect partially occluded humans in still images is proposed. An individual human is modeled as an assembly of natural body parts. Some part based SVM classifiers are trained first by using histogram of orientated gradient feature. Different from other boosting methods, region information is stored in each classifier. When detect human in crowed scene, according to the information of humans that have already been detected, the information of available regions could be obtained, when a new detection window is in process. In classifier sequence, the classifiers whose regions are available are selected for generating the final classifier. This method could achieve good performance on images and video sequences with several occlusions.

AB - In this paper, a novel method to detect partially occluded humans in still images is proposed. An individual human is modeled as an assembly of natural body parts. Some part based SVM classifiers are trained first by using histogram of orientated gradient feature. Different from other boosting methods, region information is stored in each classifier. When detect human in crowed scene, according to the information of humans that have already been detected, the information of available regions could be obtained, when a new detection window is in process. In classifier sequence, the classifiers whose regions are available are selected for generating the final classifier. This method could achieve good performance on images and video sequences with several occlusions.

UR - http://www.scopus.com/inward/record.url?scp=70349925460&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70349925460&partnerID=8YFLogxK

U2 - 10.1109/CSPA.2009.5069221

DO - 10.1109/CSPA.2009.5069221

M3 - Conference contribution

SN - 9781424441501

SP - 224

EP - 227

BT - Proceedings of 2009 5th International Colloquium on Signal Processing and Its Applications, CSPA 2009

ER -