From stixels to asteroids: towards a collision warning system using stereo vision

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

Uittreksel

This paper explores the use of stixels in a probabilistic stereo vision-based collision-warning system that can be part of an ADAS for intelligent vehicles. In most current systems, collision warnings are based on radar or on monocular vision using pattern recognition (and ultra-sound for park assist). Since detecting collisions is such a core functionality of intelligent vehicles, redundancy is key. Therefore, we explore the use of stereo vision for reliable collision prediction. Our algorithm consists of a Bayesian histogram filter that provides the probability of collision for multiple interception regions and angles towards the vehicle. This could additionally be fused with other sources of information in larger systems. Our algorithm builds upon the disparity Stixel World that has been developed for efficient automotive vision applications. Combined with image flow and uncertainty modeling, our system samples and propagates asteroids, which are dynamic particles that can be utilized for collision prediction. At best, our independent system detects all 31 simulated collisions (2 false warnings), while this setting generates 12 false warnings on the real-world data.
Originele taal-2Engels
TitelIS&T Electronic Imaging
SubtitelAutonomous Vehicles and Machines
UitgeverijSociety for Imaging Science and Technology
Aantal pagina's6
DOI's
StatusGepubliceerd - 14 jan 2019
EvenementIS&T International Symposium on Electronic Imaging 2019, Image Processing: Algorithms and Systems XVII - Burlingame, Verenigde Staten van Amerika
Duur: 13 jan 201917 jan 2019
Congresnummer: XVII
http://www.imaging.org/site/IST/IST/Conferences/EI/EI_2019/Conference/C_IPAS.aspx

Congres

CongresIS&T International Symposium on Electronic Imaging 2019, Image Processing: Algorithms and Systems XVII
Verkorte titelIPAS2019
LandVerenigde Staten van Amerika
StadBurlingame
Periode13/01/1917/01/19
Internet adres

Vingerafdruk

Asteroids
Intelligent vehicle highway systems
Stereo vision
Alarm systems
Pattern recognition
Redundancy
Radar
Ultrasonics
Uncertainty

Citeer dit

Sanberg, W., Dubbelman, G., & de With, P. (2019). From stixels to asteroids: towards a collision warning system using stereo vision. In IS&T Electronic Imaging: Autonomous Vehicles and Machines [034] Society for Imaging Science and Technology. https://doi.org/10.2352/ISSN.2470-1173.2019.15.AVM-034
Sanberg, Willem ; Dubbelman, Gijs ; de With, Peter. / From stixels to asteroids: towards a collision warning system using stereo vision. IS&T Electronic Imaging: Autonomous Vehicles and Machines. Society for Imaging Science and Technology, 2019.
@inproceedings{0f37071098fa47daad177e0ba7aa47ef,
title = "From stixels to asteroids: towards a collision warning system using stereo vision",
abstract = "This paper explores the use of stixels in a probabilistic stereo vision-based collision-warning system that can be part of an ADAS for intelligent vehicles. In most current systems, collision warnings are based on radar or on monocular vision using pattern recognition (and ultra-sound for park assist). Since detecting collisions is such a core functionality of intelligent vehicles, redundancy is key. Therefore, we explore the use of stereo vision for reliable collision prediction. Our algorithm consists of a Bayesian histogram filter that provides the probability of collision for multiple interception regions and angles towards the vehicle. This could additionally be fused with other sources of information in larger systems. Our algorithm builds upon the disparity Stixel World that has been developed for efficient automotive vision applications. Combined with image flow and uncertainty modeling, our system samples and propagates asteroids, which are dynamic particles that can be utilized for collision prediction. At best, our independent system detects all 31 simulated collisions (2 false warnings), while this setting generates 12 false warnings on the real-world data.",
author = "Willem Sanberg and Gijs Dubbelman and {de With}, Peter",
year = "2019",
month = "1",
day = "14",
doi = "10.2352/ISSN.2470-1173.2019.15.AVM-034",
language = "English",
booktitle = "IS&T Electronic Imaging",
publisher = "Society for Imaging Science and Technology",
address = "United States",

}

Sanberg, W, Dubbelman, G & de With, P 2019, From stixels to asteroids: towards a collision warning system using stereo vision. in IS&T Electronic Imaging: Autonomous Vehicles and Machines., 034, Society for Imaging Science and Technology, IS&T International Symposium on Electronic Imaging 2019, Image Processing: Algorithms and Systems XVII, Burlingame, Verenigde Staten van Amerika, 13/01/19. https://doi.org/10.2352/ISSN.2470-1173.2019.15.AVM-034

From stixels to asteroids: towards a collision warning system using stereo vision. / Sanberg, Willem; Dubbelman, Gijs; de With, Peter.

IS&T Electronic Imaging: Autonomous Vehicles and Machines. Society for Imaging Science and Technology, 2019. 034.

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

TY - GEN

T1 - From stixels to asteroids: towards a collision warning system using stereo vision

AU - Sanberg, Willem

AU - Dubbelman, Gijs

AU - de With, Peter

PY - 2019/1/14

Y1 - 2019/1/14

N2 - This paper explores the use of stixels in a probabilistic stereo vision-based collision-warning system that can be part of an ADAS for intelligent vehicles. In most current systems, collision warnings are based on radar or on monocular vision using pattern recognition (and ultra-sound for park assist). Since detecting collisions is such a core functionality of intelligent vehicles, redundancy is key. Therefore, we explore the use of stereo vision for reliable collision prediction. Our algorithm consists of a Bayesian histogram filter that provides the probability of collision for multiple interception regions and angles towards the vehicle. This could additionally be fused with other sources of information in larger systems. Our algorithm builds upon the disparity Stixel World that has been developed for efficient automotive vision applications. Combined with image flow and uncertainty modeling, our system samples and propagates asteroids, which are dynamic particles that can be utilized for collision prediction. At best, our independent system detects all 31 simulated collisions (2 false warnings), while this setting generates 12 false warnings on the real-world data.

AB - This paper explores the use of stixels in a probabilistic stereo vision-based collision-warning system that can be part of an ADAS for intelligent vehicles. In most current systems, collision warnings are based on radar or on monocular vision using pattern recognition (and ultra-sound for park assist). Since detecting collisions is such a core functionality of intelligent vehicles, redundancy is key. Therefore, we explore the use of stereo vision for reliable collision prediction. Our algorithm consists of a Bayesian histogram filter that provides the probability of collision for multiple interception regions and angles towards the vehicle. This could additionally be fused with other sources of information in larger systems. Our algorithm builds upon the disparity Stixel World that has been developed for efficient automotive vision applications. Combined with image flow and uncertainty modeling, our system samples and propagates asteroids, which are dynamic particles that can be utilized for collision prediction. At best, our independent system detects all 31 simulated collisions (2 false warnings), while this setting generates 12 false warnings on the real-world data.

U2 - 10.2352/ISSN.2470-1173.2019.15.AVM-034

DO - 10.2352/ISSN.2470-1173.2019.15.AVM-034

M3 - Conference contribution

BT - IS&T Electronic Imaging

PB - Society for Imaging Science and Technology

ER -

Sanberg W, Dubbelman G, de With P. From stixels to asteroids: towards a collision warning system using stereo vision. In IS&T Electronic Imaging: Autonomous Vehicles and Machines. Society for Imaging Science and Technology. 2019. 034 https://doi.org/10.2352/ISSN.2470-1173.2019.15.AVM-034