Development of Pedestrian Property Estimation Method Based on Deep Neural Networks Using LiDAR Sensor

Taejin Kim; Eunbi Jeong; Soyoung Iris You

doi:10.7470/jkst.2018.36.5.319

All Issue

2018 Vol.36, Issue 5 Preview Page Next Page

Article

Development of Pedestrian Property Estimation Method Based on Deep Neural Networks Using LiDAR Sensor LiDAR 센서 기반의 보행 동선 자료를 활용한 보행자 속성 추정 방법론 개발

31 October 2018. pp. 319-330

PDF XML

Abstract

As the number of accidents and crimes in Seoul metro system has recently increased, there has been growing concerns on the safety of passengers. Even though the integrated management method of metro station using the intelligent CCTV is proposed as an alternative, the cost saving and the privacy issue in analyzing and operating the image based information is critical. Recently, many studies regarding a variety types of sensors have been throughly investigated. Among all, LiDAR sensor has excellent capability of recognizing and tracking objects, and also has the advantages of realizing information acquisition and analysis in real time. In this study, the characteristics of pedestrians are extracted from individual pedestrians’ trajectory using LiDAR sensor, and analyzed. With the proposed Deep Neural Network (DNN) algorithm in this study, pedestrian property can be predictable from each individual trajectory. As a result, we found a great potential to monitor and manage effectively and seek the certain situation reflecting the behavior of pedestrian’s characteristic with environmental elements in a metro station.

Keywords

deep neural network

LiDAR sensor

pedestrian characteristics

pedestrian trajectory

pedestrian

최근 도시철도 내 사건사고 및 범죄 수가 급격히 증가하면서 승객의 안전에 대한 관심은 점차 늘어나고 있다. 지능형 CCTV를 활용한 도시철도역사 통합 관리 방안 등 다양한 전략이 제시되고 있으나, 비용 절감 및 개인사생활 침해 등은 여전히 이슈가 되고 있다. 최근 기술 개발로 인한 다양한 센서에 대한 연구가 진행되고 있으며, 그 중에서도 LiDAR 센서는 사물 인식, 추적 그리고 실시간 정보 처리 등 장점이 많다. 본 연구에서는 이러한 LiDAR 센서를 활용하여, 개별 보행자의 동선 정보로부터 개인 속성을 추출하였으며, 제안된 신경망 알고리즘 기반 모형을 통해 보행자의 속성을 추정하였다. 분석결과를 통해 도시철도역사 환경 요소와 개인 속성을 고려한 행태 정보를 반영한 보행 흐름에 대한 활용성 및 도시철도 역사에 대한 효율적인 유지관리 가능성을 제시하였다.

키워드

심층신경망

라이다센서

보행자 속성

보행 동선

보행자

References

Alfonso G. B., Gabriel V., Juan F. D. P., Juan M. C. (2018), A Multi-agent System for the Classification of Gender and Age from Images, Computer Vision and Image Understanding, In press, Available online 6,

10.1016/ j.cviu.2018.01.012

Boltes M., Seyfried A. (2013), Collecting Pedestrian Trajectories, Neurocomputing, 16, 127-133.

10.1016/j.neucom.2012.01.036

Borges P. (2013), Pedestrian Detection Based on Blob Motion Statistics, IEEE Transactions on Circuits and Systems for Video Technology, 224-235.

10.1109/TCSVT.2012.2203217

Downs R., Stea D. (2011), Cognitive Maps and Spatial Behaviour: Process and Products, John Wiley & Sons Ltd, 312-317.

Hossein M., Saeed M. (2017), Gender Dictionary Learning for Gender Classification, J. Vis. Commun. Image R. 42, 1-13.

10.1016/j.jvcir.2016.11.002

Jagtap J., Kokare M. (2016), Human age Classification Using Facial Skin Aging Features and Artificial Neural Network, Cognitive Systems Research, 40, 116-128.

10.1016/j.cogsys.2016.05.002

Jeong E., You S. (2017), An Algorithm of Identifying Roaming Pedestrians' Trajectories using LiDAR Sensor, Journal of the Korea Institute of Intelligent Transport Systems, 1-15.

Jeong E., You S., Lee J., Kim K. (2017), A Comprehensive Framework for Estimating Pedestrian OD Matrix Using Spatial Information and Integrated Smart Card Data, J. Korean Soc. Transp., 35(5), Korean Society of Transportation, 409-422.

Kang H., Park J., Song J., Yoon B. (2016), CCTV Based Gender Classification Using a Convolutional Neural Network, Journal of Korea Multimedia Society, 1943-1950.

Kim Y., Kim C., Cho I., Kim A., Nam S. (2007), The Effect of Spatial Layout on Space Use Pattern in Subway Station, Conference of the Korean Society for Railway, 7-18.

Lee J., Kim T., You S. (2015), Improvement of Pedestrian Convenience and Mobility by Applying the Walking Guidance System in Subway Stations, J. Korean Soc. Transp., 33(2), Korean Society of Transportation, 204-213.

Lee J., Kim T., You S. (2015), The Implementation of Walking Guidance System in Subway Stations and its Effect on Decreasing Pedestrian Delay, The 72th Conference of KST, Korean Society of Transportation, 216-219.

Xue H., Liu Y., Cai D., He X. (2016), Tracking People in RGBD Videos Using Deep Learning and Motion Clues, 5, 70-76.

You S., Jung R., Chung J. (2016), An Analysis on Evacuation Scenario at Metro-stations using Pedestrian Movement- based Simulation Model, Journal of the Korea Institute of Intelligent Transport Systems, 36-49.

Yun Y., Kim Y., Park Y. (2005), The Effect of Metric Distance and Spatial Depth on Spatial Cognition, Journal of the Architectural Institute of Korea Planning & Design, 77-84.

Information

Publisher :Korean Society of Transportation
Publisher(Ko) :대한교통학회
Journal Title :Journal of Korean Society of Transportation
Journal Title(Ko) :대한교통학회지
Volume : 36
No :5
Pages :319-330
Received Date : 2018-06-04
Revised Date : 2018-06-25
Accepted Date : 2018-10-10
DOI :https://doi.org/10.7470/jkst.2018.36.5.319

Journal of Korean Society of Transportation ISSN:1229-1366(Print) 2234-4217(Online) 대한교통학회지

All Issue