Vysoké učení technické v Brně, FEKT

You are here

Projekt

Project designation:
BAD

Principal investigator:
Ing. Petr Honzík, PhD.

Project duration:
2009-2012

Provider:
GACR

Car transport safety

The project has two main areas of research: evaluation of the potential effect of vehicular communications on road traffic and indirect detection of driver’s drowsiness from the vehicle behavior. According to the project goals, a traffic and communication simulator enabling direct simulation of intelligent transportation system scenarios was proposed and developed. Extensive measurements of...

Description

The project has two main areas of research: evaluation of the potential effect of vehicular communications on road traffic and indirect detection of driver’s drowsiness from the vehicle behavior. According to the project goals, a traffic and communication simulator enabling direct simulation of intelligent transportation system scenarios was proposed and developed. Extensive measurements of communication range and reliability of wireless vehicular communications under real driving conditions were performed. A model of information dissemination in vehicular communication networks was proposed and implemented in the developed simulator. The communication parameters of the simulation model are adapted from results of the performed field measurements. The potential effects of vehicular communications on road traffic were evaluated in the developed simulator. The performed simulation experiments indicate that vehicular communications may bring significant benefits by utilizing the available road network capacity and reducing congestion. For the cars equipped with vehicular communication and smart navigation system the average travel time can be significantly reduced. For the second task, a telematics system for synchronized data acquisition from the engine communication bus and the video stream of the driver was developed and successfully used for measurements. An original methodology for directly assessing the driver’s drowsiness level was proposed. The methodology is defined as a fusion of objective (features extracted from the video stream, psycho-diagnostic tests) and subjective (driver’s personal feeling) measurements. A series of test drives under real traffic conditions was performed utilizing the developed methods and systems. The data collected during the test drives were made publicly available in an online data archive. The obtained data were used to solve the task of indirect detection of the driver’s drowsiness from the vehicle behavior (pattern recognition, parametric model of a driver, trajectory analysis). A set of ten relevant features based on steering wheel motion enabling accurate recognition of the driver’s drowsiness levels was identified. Statistical characteristics indicating non-standard driver behavior based on vehicle trajectory were found. For the classification task several original algorithms were developed and successfully published, e.g. filter feature selection by IWA, new population based optimization method PSV or new algorithm for multi-class AUC calculation.

Project goals summary

EVALUATION OF VEHICULAR COMMUNICATION TECHNOLOGIES EFFECTS ON TRAFFIC: The goal is to estimate the potential effect of using wireless vehicular communication technologies and intelligent transportation system applications on road traffic. This is achieved by measurements of vehicular communication parameters under real world conditions and simulation experiments in a specialized simulator TRASI developed for this purpose.

PUBLIC ARCHIVE OF DATA FROM TEST DRIVES: The goal is to create a publicly available archive of data collected during the operation of a car under real traffic conditions. Each record in the archive includes 1) data collected from the engine communication bus and 2) measurement of the driver’s drowsiness estimated by direct monitoring methods (measurement on the driver).

INDIRECT DRIVER MONITORING: The goal is to develop methods and algorithms for detection of increase in the driver’s drowsiness based on the changes in the vehicle’s behavior – Indirect Driver Monitoring.

 

left
right

Publications

Journal papers (Jimp)

RONZHINA, M.; JANOUŠEK, O.; KOLÁŘOVÁ, J.; NOVÁKOVÁ, M.; HONZÍK, P.; PROVAZNÍK, I. Sleep Scoring Using Artificial Neural Networks, SLEEP MEDICINE REVIEWS, 2012, vol. 16, p. 251-263

HONZÍK, P.; KŘIVAN, L.; LOKAJ, P.; LÁBROVÁ, R.; NOVÁKOVÁ, Z.; FIŠER, B.; HONZÍKOVÁ, N. Logit and Fuzzy Models in Data Analysis: Estimation of Risk in Cardiac Patients. Physiological Research, 2010, roč. 59, č. Suppl. 1, s. 89-96. ISSN: 0862- 8408.1, s. 1-13. ISSN: 1087- 0792.

Conference papers indexed in Conference Proceedings Citation Index, Thomson Reuters (D)

HYNČICA, O.; HONZÍK, P.; KUČERA, P.; PAVLATA, K. Urban Vehicle-toInfrastructure Wireless Communications Range Evaluation, IEEE ITSC, 2012, p. 915-920

RONZHINA, M.; BUBNÍK, K.; GAJDOŠ, M.; KOLÁŘOVÁ, J.; HONZÍK, P.; PROVAZNÍK, I. Use of EEG for Validation of Flicker- Fusion Test. In 4TH INTERNATIONAL SYMPOSIUM ON APPLIED SCIENCES IN BIOMEDICAL AND COMMUNICATION TECHNOLOGIES - ISABEL. Barcelona, Spain: 2011.

HONZÍK, P.; RÁŠO, P.; KUČERA, P.; PAVLATA, K.; HAUPT, D. Vehicle Data Acquisition and Analysis for Detection of the Driver' s Drowsiness. In The proceedings of the 17th International Conference on Soft Computing. 2011. s. 1-6. ISBN: 978-80-214-4302- 0.

HORÁK, K.; HONZÍK, P.; KUČERA, P. On Image Segmentation Techniques for Driver Inattention Systems. In The proceedings of the 17th International Conference on Soft Computing. Brno, Czech Republic: Department of Telecommunications, Tribun EU s.r.o., 2011. s. 1-6. ISBN: 978-80-214-4120- 0.

BURIAN, F.; FLORIÁN, T.; ŽALUD, L. The Identification of drivers behaviour through the use of GPS and odometry. In Proceedings of Mendel 2011. Mendel Journal series. Brno, CZ. 2011. s. 492-497. ISBN: 978-80-214-4302- 0.

HYNČICA, T.; HYNČICA, O.; KUČERA, L. SIMULATION ANALYSIS OF URBAN TRAFFIC FLOWS WITH ALTERNATE ROUTES. In MENDEL 2011 17th Intermational Conference on Soft Computing. 1. Brno, Nosová Hana, Koperníkova 4362/21, Brno, 2011. p. 282 - 288. ISBN 978-80-214-4302-0.

HONZÍK, P.; KUČERA, P.; HYNČICA, O.; HAUPT, D. New Feature Selection Method for Multi-class Data: Iteratively Weighted AUC (IWA). In The 6th IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications. 2011. p. 336 - 340. ISBN 978-1-4577-1423-8.

KUČERA, P.; HYNČICA, O.; HONZÍK, P.; PAVLATA, K.; HORÁK, K. On Vehicle Data Acquisition System. In IEEE ITSC 2011. Washington, USA. 2011. p. 1446 - 1451. ISBN 978-1-4577-2196-0.

ŠTOHL, R.; HAVLÍKOVÁ, M.; ŠEDIVÁ, S. Computer Analysis of Driver Trajectory. In Proceedings of the 15th WSEAS International Conference on Computers. Greece: WSEAS Press, 2011. s. 135-140. ISBN: 978-1-61804-019- 0.

HORÁK, K.; KALOVÁ, I. Eyes Detection and Tracking for Monitoring Driver Vigilance. In The proceedings of the 33rd International Conference on Telecommunication and Signal Processing. H- 1055 Budapest, Szent István krt. 7.: Asszisztencia Szervezo Kft., 2010. s.204-208. ISBN: 978-963-88981-0- 4.

POLÁCH, P.; MILLION, P.; HONZÍK, P.; POHL, J. Petri Nets in Expert Systems. In Proceedings of the 16th International Conference on Soft Computing. 1. Brno: 2010. s. 216-222. ISBN: 978-960-474-178- 6.

HONZÍK, P.; HYNČICA, O.; KUČERA, P.; JIRSÍK, V. OPTIMIZATION CHALLENGE: AREA UNDER RECEIVER OPERATING CHARACTERISTIC AS A LOSS FUNCTION IN GENERALIZED LINEAR MODELS. In MENDEL 2010. 1. Brno, Czech Republic: Brno University of Technology, 2010. s. 29-34. ISBN: 978-80-214-4120- 0.

POHL, J.; HONZÍK, P.; JIRSÍK, V.; POLÁCH, P. Filtration of samples for secondary surveillance radar. In MENDEL 2010. 2010. s. 260-265. ISBN: 978-80-214-4120- 0.

HORÁK, K.; KALOVÁ, I. Image Acquisition and Processing for Monitoring Driver Vigilance. In The proceedings of the 16th International Conference on Soft Computing. Brno: Brno University of Technology, 2010. s. 517-523. ISBN: 978-80-214-4120- 0.

HONZÍK, P.; KUČERA, P.; HYNČICA, O.; JIRSÍK, V. Novel Method for Evaluation of Multi- class Area Under Receiver Operating Characteristic. In 5th International Conference on Soft Computing, Computing with Words and Perceptions in System Analysis, Decision and Control. North Cyprus: IEEE, 2009. s. 179-182. ISBN: 978-1-4244-3428- 2.

HONZÍK, P.; KUČERA, P.; HYNČICA, O.; JIRSÍK, V. AREA UNDER THE RECEIVER OPERATING CHARACTERISTIC AS A SORTING PROBLEM. In MENDELL 2009, 15th International Conference on Soft Computing. Brno, Czech Republic: 2009. s. 277-282. ISBN: 978-80-214-3884- 2.

POLÁCH, P.; POHL, J. Simulation Platform for Artificial Intelligence and Soft Computing. In Mendel Journal series. 1. Brno: 2009. s. 228-233. ISBN: 978-80-214-3884- 2.

HORÁK, K.; RICHTER, M.; KALOVÁ, I. Human Eyes Localization for Driver Inattention Monitoring System. In Book the Proceedings of MENDEL 2009. Mendel Journal Series 2009. Brno, Czech Republic: Brno University of Technology/ Institution of Automation and Computer Science, 2009. s. 283-288. ISBN: 978-80-214-3884- 2. ISSN: 1803- 3814.

POHL, J.; POLÁCH, P. Stochastic learning algorithm. In Mendel Journal series. 2009. s. 166-170. ISBN: 978-80-214-3884- 2.

Project results

Principal investigator

  • Ing. Petr Honzík, PhD.

Department of Control and Instrumentation, Brno University of TechnologyTechnická 3082/12616 00 BrnoCzech RepublicCIN: 00216305, tel.: +420 5 4114-6411fax+420 5 4114-6451