Title: Symmetrical judgement area reduction and ECoHOG feature descriptor for pedestrian detection

Authors: Hirokatsu Kataoka; Yoshimitsu Aoki; Yasuhiro Matsui

Addresses: Graduate School of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku-ku Yokohama, Kanagawa, 223-8522, Japan ' Graduate School of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku-ku Yokohama, Kanagawa, 223-8522, Japan ' National Traffic Safety and Environment Laboratory, 7-42-27 Jindaiji-Higashi-machi, Chofu, Tokyo 182-0012, Japan

Abstract: In this study, a method to detect pedestrians using an in-vehicle camera is presented. We improved the technology in detecting pedestrians with highly accurate images using a monocular camera. We were able to predict pedestrians activities by monitoring them, and we developed an algorithm to recognise pedestrians and their movements more accurately. For the feature descriptor, we found that an Extended Co-occurrence Histogram of Oriented Gradients (ECoHOG) was the best in decreasing both the undetectable and the excessive detectable ratio. Thus, the use of the new method by images captured on the real road was validated.

Keywords: active safety; pedestrian detection; monocular cameras; symmetrical judgement area reduction; ECoHOG; extended co-occurrence histogram; oriented gradients; pedestrians; in-vehicle cameras; pedestrian monitoring; pedestrian recognition; pedestrian safety; vehicle safety.

DOI: 10.1504/IJVS.2012.048548

International Journal of Vehicle Safety, 2012 Vol.6 No.1, pp.48 - 60

Published online: 31 Dec 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article