Home > Archive > 2016 > Volume 6 Number 2 (Apr. 2016) >
IJMLC 2016 Vol.6(2): 87-91 ISSN: 2010-3700
DOI: 10.18178/ijmlc.2016.6.2.578

Pedestrian Detection Using Objectness Information

Weidong Yao, Xiaohui Chen, Li Chen, and Weidong Wang

Abstract—In recent ten years, the research works about pedestrian detection using various methods are coming forth one after another, pushing the state of the art in this area. Diverse features and excellent classification technique are the key to get prime performance in these research activities. We are now motived to present a more effective approach from these two aspects. Be inspired by the research achievement in objectness estimation field, our pedestrian detection method integrates boosted classifier with multiple objectness features, including salient feature and edgebox feature. At the same time, we improve classifier structure to achieve a better performance, the influence of classifier threshold on the prediction is also analyzed in this paper. As a practice part, we extract data from several exist datasets, and append traffic scene images of our city to form a new dataset for studying about the generalization ability of pedestrian detection. Our novel approach for pedestrian detection demonstrates significant performance advantages on precision and generalization ability by a series of experiments.

Index Terms—Pedestrian detection, aggregated channel features , objectness, real adaboost algorithm.

The authors are with the Dept. of Electronic Engineering and Information Science, University of Science and Technology of China, Hefei, China (e-mail: {yao1990, clbyx}@mail.ustc.edu.cn, {cxh, wdwang}@ustc.edu.cn).


Cite: Weidong Yao, Xiaohui Chen, Li Chen, and Weidong Wang, "Pedestrian Detection Using Objectness Information," International Journal of Machine Learning and Computing vol.6, no. 2, pp. 87-91, 2016.

General Information

  • E-ISSN: 2972-368X
  • Abbreviated Title: Int. J. Mach. Learn.
  • Frequency: Quaterly
  • DOI: 10.18178/IJML
  • Editor-in-Chief: Dr. Lin Huang
  • Executive Editor:  Ms. Cherry L. Chen
  • Abstracing/Indexing: Inspec (IET), Google Scholar, Crossref, ProQuest, Electronic Journals LibraryCNKI.
  • E-mail: ijml@ejournal.net

Article Metrics in Dimensions