Fast human detection using selective block-based HOG-LBP

Won Jae Park, Dae Hwan Kim, Suryanto, Chun Gi Lyuh, Tae Moon Roh, Sung-Jea Ko

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

15 Citations (Scopus)

Abstract

We propose a speed up method for the Histograms of Oriented Gradients - Local Binary Pattern (HOG-LBP) based pedestrian detector. Our method is based on the two-stage cascade structure. In the first stage evaluation, instead of extracting the features from all the region inside the detection window like in the conventional method, we extract the features from the regions which best characterize the pedestrian only. By reducing the features to be evaluated, each candidate is evaluated faster. To determine which regions are best for characterizing the pedestrian, we train the AdaBoost classifier to select the blocks whose Support Vector Machine responses of the pedestrian samples are most different from the non-pedestrians. In the second stage, we simply use the conventional HOG-LBP classifier to reevaluate the candidates which pass the first stage evaluation. Experimental results show that the detection algorithm is about three times faster than the conventional HOG-LBP SVM algorithm.

Original languageEnglish
Title of host publicationProceedings - International Conference on Image Processing, ICIP
Pages601-604
Number of pages4
DOIs
Publication statusPublished - 2012 Dec 1
Event2012 19th IEEE International Conference on Image Processing, ICIP 2012 - Lake Buena Vista, FL, United States
Duration: 2012 Sep 302012 Oct 3

Other

Other2012 19th IEEE International Conference on Image Processing, ICIP 2012
CountryUnited States
CityLake Buena Vista, FL
Period12/9/3012/10/3

Fingerprint

Classifiers
Adaptive boosting
Support vector machines
Detectors

Keywords

  • Block-Based
  • Cascade
  • Fast
  • HOG-LBP Feature
  • Human Detection

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems

Cite this

Park, W. J., Kim, D. H., Suryanto, Lyuh, C. G., Roh, T. M., & Ko, S-J. (2012). Fast human detection using selective block-based HOG-LBP. In Proceedings - International Conference on Image Processing, ICIP (pp. 601-604). [6466931] https://doi.org/10.1109/ICIP.2012.6466931

Fast human detection using selective block-based HOG-LBP. / Park, Won Jae; Kim, Dae Hwan; Suryanto; Lyuh, Chun Gi; Roh, Tae Moon; Ko, Sung-Jea.

Proceedings - International Conference on Image Processing, ICIP. 2012. p. 601-604 6466931.

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

Park, WJ, Kim, DH, Suryanto, Lyuh, CG, Roh, TM & Ko, S-J 2012, Fast human detection using selective block-based HOG-LBP. in Proceedings - International Conference on Image Processing, ICIP., 6466931, pp. 601-604, 2012 19th IEEE International Conference on Image Processing, ICIP 2012, Lake Buena Vista, FL, United States, 12/9/30. https://doi.org/10.1109/ICIP.2012.6466931
Park WJ, Kim DH, Suryanto, Lyuh CG, Roh TM, Ko S-J. Fast human detection using selective block-based HOG-LBP. In Proceedings - International Conference on Image Processing, ICIP. 2012. p. 601-604. 6466931 https://doi.org/10.1109/ICIP.2012.6466931
Park, Won Jae ; Kim, Dae Hwan ; Suryanto ; Lyuh, Chun Gi ; Roh, Tae Moon ; Ko, Sung-Jea. / Fast human detection using selective block-based HOG-LBP. Proceedings - International Conference on Image Processing, ICIP. 2012. pp. 601-604
@inproceedings{3427045599ca4ea3bbc639a17956dba2,
title = "Fast human detection using selective block-based HOG-LBP",
abstract = "We propose a speed up method for the Histograms of Oriented Gradients - Local Binary Pattern (HOG-LBP) based pedestrian detector. Our method is based on the two-stage cascade structure. In the first stage evaluation, instead of extracting the features from all the region inside the detection window like in the conventional method, we extract the features from the regions which best characterize the pedestrian only. By reducing the features to be evaluated, each candidate is evaluated faster. To determine which regions are best for characterizing the pedestrian, we train the AdaBoost classifier to select the blocks whose Support Vector Machine responses of the pedestrian samples are most different from the non-pedestrians. In the second stage, we simply use the conventional HOG-LBP classifier to reevaluate the candidates which pass the first stage evaluation. Experimental results show that the detection algorithm is about three times faster than the conventional HOG-LBP SVM algorithm.",
keywords = "Block-Based, Cascade, Fast, HOG-LBP Feature, Human Detection",
author = "Park, {Won Jae} and Kim, {Dae Hwan} and Suryanto and Lyuh, {Chun Gi} and Roh, {Tae Moon} and Sung-Jea Ko",
year = "2012",
month = "12",
day = "1",
doi = "10.1109/ICIP.2012.6466931",
language = "English",
isbn = "9781467325332",
pages = "601--604",
booktitle = "Proceedings - International Conference on Image Processing, ICIP",

}

TY - GEN

T1 - Fast human detection using selective block-based HOG-LBP

AU - Park, Won Jae

AU - Kim, Dae Hwan

AU - Suryanto,

AU - Lyuh, Chun Gi

AU - Roh, Tae Moon

AU - Ko, Sung-Jea

PY - 2012/12/1

Y1 - 2012/12/1

N2 - We propose a speed up method for the Histograms of Oriented Gradients - Local Binary Pattern (HOG-LBP) based pedestrian detector. Our method is based on the two-stage cascade structure. In the first stage evaluation, instead of extracting the features from all the region inside the detection window like in the conventional method, we extract the features from the regions which best characterize the pedestrian only. By reducing the features to be evaluated, each candidate is evaluated faster. To determine which regions are best for characterizing the pedestrian, we train the AdaBoost classifier to select the blocks whose Support Vector Machine responses of the pedestrian samples are most different from the non-pedestrians. In the second stage, we simply use the conventional HOG-LBP classifier to reevaluate the candidates which pass the first stage evaluation. Experimental results show that the detection algorithm is about three times faster than the conventional HOG-LBP SVM algorithm.

AB - We propose a speed up method for the Histograms of Oriented Gradients - Local Binary Pattern (HOG-LBP) based pedestrian detector. Our method is based on the two-stage cascade structure. In the first stage evaluation, instead of extracting the features from all the region inside the detection window like in the conventional method, we extract the features from the regions which best characterize the pedestrian only. By reducing the features to be evaluated, each candidate is evaluated faster. To determine which regions are best for characterizing the pedestrian, we train the AdaBoost classifier to select the blocks whose Support Vector Machine responses of the pedestrian samples are most different from the non-pedestrians. In the second stage, we simply use the conventional HOG-LBP classifier to reevaluate the candidates which pass the first stage evaluation. Experimental results show that the detection algorithm is about three times faster than the conventional HOG-LBP SVM algorithm.

KW - Block-Based

KW - Cascade

KW - Fast

KW - HOG-LBP Feature

KW - Human Detection

UR - http://www.scopus.com/inward/record.url?scp=84875848953&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84875848953&partnerID=8YFLogxK

U2 - 10.1109/ICIP.2012.6466931

DO - 10.1109/ICIP.2012.6466931

M3 - Conference contribution

SN - 9781467325332

SP - 601

EP - 604

BT - Proceedings - International Conference on Image Processing, ICIP

ER -