A Framework For Automated Analysis of Surrogate Measures of Safety from Video using Deep Learning Techniques

Morten Bornø Jensen, Martin Ahrnbom, Maarten Kruithof, Kalle Åström, Mikael Nilsson, Håkan Ardö, Aliaksei Laureshyn, Carl Johnsson, Thomas B. Moeslund

Research output: Contribution to journalConference article in JournalResearchpeer-review

Abstract

Traffic surveillance and monitoring are gaining a lot of attention as a result of an
increase of vehicles on the road and a desire to minimize accidents. In order to minimize
accidents and near-accidents, it is important to be able to judge the safety of a
traffic environment. It is possible to perform traffic analysis using large quantities
of video data. Computer vision is a great tool for reducing the data, so that only sequences
of interest are further analyzed. In this paper, we propose a cross-disciplinary
framework for performing automated traffic analysis, from both a computer vision researcher’s
and traffic researcher’s point-of-view. Furthermore, we present STRUDL,
an open-source implementation of this framework, that computes trajectories of road
users, which we use to automatically find sequences containing critical events of
vehicles and vulnerable road users in an traffic intersection, which is an otherwise
time-consuming task.
Keywords: Computer vision, data reduction, computer aided analysis, deep learning,
surveillance, tracking, detection, traffic analysis
Original languageEnglish
JournalTransportation Research Board. Annual Meeting Proceedings
Pages (from-to)281-306
ISSNx000-0023
Publication statusPublished - Jan 2019
EventTransportation Research Board (TRB) 98th Annual Meeting - Walter E. Washington Convention Center, Washington, D.C, United States
Duration: 13 Jan 201917 Jan 2019
http://www.trb.org/AnnualMeeting/AnnualMeeting.aspx

Conference

ConferenceTransportation Research Board (TRB) 98th Annual Meeting
LocationWalter E. Washington Convention Center
CountryUnited States
CityWashington, D.C
Period13/01/201917/01/2019
Internet address

Fingerprint

Computer vision
Accidents
Computer aided analysis
Data reduction
Trajectories
Monitoring
Deep learning

Cite this

Jensen, Morten Bornø ; Ahrnbom, Martin ; Kruithof, Maarten ; Åström, Kalle ; Nilsson, Mikael ; Ardö, Håkan ; Laureshyn, Aliaksei ; Johnsson, Carl ; Moeslund, Thomas B. / A Framework For Automated Analysis of Surrogate Measures of Safety from Video using Deep Learning Techniques. In: Transportation Research Board. Annual Meeting Proceedings. 2019 ; pp. 281-306.
@inproceedings{ad7594a09d004013a9b5b4e4938499b6,
title = "A Framework For Automated Analysis of Surrogate Measures of Safety from Video using Deep Learning Techniques",
abstract = "Traffic surveillance and monitoring are gaining a lot of attention as a result of anincrease of vehicles on the road and a desire to minimize accidents. In order to minimizeaccidents and near-accidents, it is important to be able to judge the safety of atraffic environment. It is possible to perform traffic analysis using large quantitiesof video data. Computer vision is a great tool for reducing the data, so that only sequencesof interest are further analyzed. In this paper, we propose a cross-disciplinaryframework for performing automated traffic analysis, from both a computer vision researcher’sand traffic researcher’s point-of-view. Furthermore, we present STRUDL,an open-source implementation of this framework, that computes trajectories of roadusers, which we use to automatically find sequences containing critical events ofvehicles and vulnerable road users in an traffic intersection, which is an otherwisetime-consuming task.Keywords: Computer vision, data reduction, computer aided analysis, deep learning,surveillance, tracking, detection, traffic analysis",
author = "Jensen, {Morten Born{\o}} and Martin Ahrnbom and Maarten Kruithof and Kalle {\AA}str{\"o}m and Mikael Nilsson and H{\aa}kan Ard{\"o} and Aliaksei Laureshyn and Carl Johnsson and Moeslund, {Thomas B.}",
year = "2019",
month = "1",
language = "English",
pages = "281--306",
journal = "Transportation Research Board. Annual Meeting Proceedings",
issn = "x000-0023",
publisher = "Transportation Research Board National Cooperative Highway Research Program",
note = "null ; Conference date: 13-01-2019 Through 17-01-2019",
url = "http://www.trb.org/AnnualMeeting/AnnualMeeting.aspx",

}

A Framework For Automated Analysis of Surrogate Measures of Safety from Video using Deep Learning Techniques. / Jensen, Morten Bornø; Ahrnbom, Martin; Kruithof, Maarten; Åström, Kalle; Nilsson, Mikael; Ardö, Håkan; Laureshyn, Aliaksei; Johnsson, Carl; Moeslund, Thomas B.

In: Transportation Research Board. Annual Meeting Proceedings, 01.2019, p. 281-306.

Research output: Contribution to journalConference article in JournalResearchpeer-review

TY - GEN

T1 - A Framework For Automated Analysis of Surrogate Measures of Safety from Video using Deep Learning Techniques

AU - Jensen, Morten Bornø

AU - Ahrnbom, Martin

AU - Kruithof, Maarten

AU - Åström, Kalle

AU - Nilsson, Mikael

AU - Ardö, Håkan

AU - Laureshyn, Aliaksei

AU - Johnsson, Carl

AU - Moeslund, Thomas B.

PY - 2019/1

Y1 - 2019/1

N2 - Traffic surveillance and monitoring are gaining a lot of attention as a result of anincrease of vehicles on the road and a desire to minimize accidents. In order to minimizeaccidents and near-accidents, it is important to be able to judge the safety of atraffic environment. It is possible to perform traffic analysis using large quantitiesof video data. Computer vision is a great tool for reducing the data, so that only sequencesof interest are further analyzed. In this paper, we propose a cross-disciplinaryframework for performing automated traffic analysis, from both a computer vision researcher’sand traffic researcher’s point-of-view. Furthermore, we present STRUDL,an open-source implementation of this framework, that computes trajectories of roadusers, which we use to automatically find sequences containing critical events ofvehicles and vulnerable road users in an traffic intersection, which is an otherwisetime-consuming task.Keywords: Computer vision, data reduction, computer aided analysis, deep learning,surveillance, tracking, detection, traffic analysis

AB - Traffic surveillance and monitoring are gaining a lot of attention as a result of anincrease of vehicles on the road and a desire to minimize accidents. In order to minimizeaccidents and near-accidents, it is important to be able to judge the safety of atraffic environment. It is possible to perform traffic analysis using large quantitiesof video data. Computer vision is a great tool for reducing the data, so that only sequencesof interest are further analyzed. In this paper, we propose a cross-disciplinaryframework for performing automated traffic analysis, from both a computer vision researcher’sand traffic researcher’s point-of-view. Furthermore, we present STRUDL,an open-source implementation of this framework, that computes trajectories of roadusers, which we use to automatically find sequences containing critical events ofvehicles and vulnerable road users in an traffic intersection, which is an otherwisetime-consuming task.Keywords: Computer vision, data reduction, computer aided analysis, deep learning,surveillance, tracking, detection, traffic analysis

UR - http://amonline.trb.org/search-results/site-search-7.291302?tag_co=&tag_lo=&ev=&tr=&tag_or=&tag_ta=&ty=&pe=&q=Moeslund&tag_su=&pn=&tag_pe=&qr=1

M3 - Conference article in Journal

SP - 281

EP - 306

JO - Transportation Research Board. Annual Meeting Proceedings

JF - Transportation Research Board. Annual Meeting Proceedings

SN - x000-0023

Y2 - 13 January 2019 through 17 January 2019

ER -