Watch the traffic signal: a signal state extraction system for naturalistic driving videos

Dongxi Zheng
Madhav V. Chitturi
David A. Noyce
John Lee

Naturalistic driving studies (NDS) are revolutionizing road safety research. Naturalistic driving data provide a new window into driver behavior that promises a deeper understanding than was ever possible with crash data, roadside observations, or driving simulator experiments. NDS collect extensive vehicle network data, vehicle internal videos (e.g., drivers and passengers), and vehicle external videos (e.g., forward roadway). The video record of the driver and surrounding road situation often provide a more revealing account of driver behavior. Data size becomes a double edged sword for most NDS. Fully or partially automated procedures are needed for data reduction. Both intentional and unintentional violations of traffic signals can cause severe consequences, like fatal right-angle crashes. Knowing the signal state when the driver navigated through the intersection is the first step towards judging the driver’s compliance and analyzing deeper human factors issues. A system that codes traffic signal state from georeferenced front-view videos was developed for use with the Second Strategic Highway Research Program (SHRP2) NDS data. GPS coordinates and a free online map database are used to identify candidate frames from lengthy videos for computer vision processing. The computer vision algorithm uses color histograms and shape matching to detect traffic signals. Vehicle movement is used to select, from multiple signals the one corresponding to the vehicle’s traffic lane. Temporal relationship between frames is employed for the purpose of refining detection results. Experiments were conducted on daytime videos and showed reasonable performance given the severe challenges posed by the SHRP2 data. Misclassifications were primarily due to other vehicles’ taillights, reddish yellow signals, yellowish red signals, green traffic signs, etc. Misses were due to distant frames with very few pixels corresponding to the signals and reduced color conspicuity compared to background scenes (e.g., green signals and the sky look similar). The system is an insightful first step towards using computer vision techniques to support signal state coding of the large volumes of naturalistic driving video data. Challenges revealed by the experimental results are valuable knowledge for future improvement.

MEET US


7
Dec

Shipping and the environment – research meets reality

Centre for Transport Studies (CTS) in co-operation with Ports of Stockholm invite you to the seminar Shipping and the environment – research meets reality.

LATEST NEWS


2017-10-26

Simulation of cut-in by manually driven vehicles in platooning scenarios

A study in a VTI-driving simulator has showed that a platoon will be able to handle a cut in from a manually driven car. The results of this study have recently been presented at two conferences in Japan.


2017-10-16

ERPUG Forum

The five-year anniversary of European Road Profile Users' Group (ERPUG) Forum will take place at Ramboll head quarter, Copenhagen, Denmark October 19-20, 2017.


2017-09-29

Self-driving buses in Sweden next year?

A self-driving, fossil-free bus. This idea might become reality through a forthcoming collaborative project involving the Swedish National Road and Transport Research Institute (VTI), Linköping University and several other participants. The project group aim...


2017-09-29

New climate-proof solutions for hard surfaces in cities

High-density road infrastructure that emphasise maximum durability and minimum maintenance create inflexible systems, which put increased stress on urban trees and lead to increased risk for flooding. Over the past five years, the ‘Climate-proof solutions for...


2017-09-29

VTI is preparing for automated vehicles

Automation of traffic systems will lead to major changes. The European Union’s (EU) CoEXist research project began in June 2017 with the aim of preparing cities and road operators for the introduction of self-driving vehicles. The Swedish National Road and...


2017-07-05

Vehicle Driver Monitoring: sleepiness and cognitive load

To prevent road crashes it is important to understand driver related contributing factors, which have been suggested to be the critical reason in 94 per cent of crashes. The overall aim of the project Vehicle Driver Monitoring has been to advance the...