Extending the stixel world with online self-supervised color modeling for road-versus-obstacle segmentation

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

12 Citations (Scopus)
88 Downloads (Pure)

Abstract

This work concentrates on vision processing for ADAS and intelligent vehicle applications. We propose a color extension to the disparity-based Stixel World method, so that the road can be robustly distinguished from obstacles with respect to erroneous disparity measurements. Our extension learns color appearance models for road and obstacle classes in an online and self-supervised fashion. The algorithm is tightly integrated within the core of the optimization process of the original Stixel World, allowing for strong fusion of the disparity and color signals. We perform an extensive evaluation, including different self-supervised learning strategies and different color models. Our newly recorded, publicly available data set is intentionally focused on challenging traffic scenes with many low-texture regions, causing numerous disparity artifacts. In this evaluation, we increase the F-score of the drivable distance from 0.86 to 0.97, compared to a tuned version of the state-of-the-art baseline method. This clearly shows that our color extension increases the robustness of the Stixel World, by reducing the number of falsely detected obstacles while not deteriorating the detection of true obstacles.
Original languageEnglish
Title of host publicationProceedings of the 17th International IEEE Conference on Intelligent Transportation Systems (ITSC), 8-10 October 2014, Qingdao, China
Place of PublicationPiscataway
PublisherInstitute of Electrical and Electronics Engineers
Pages1400-1407
ISBN (Print)978-1-4799-6077-4
DOIs
Publication statusPublished - 2014

Fingerprint

Color
Intelligent vehicle highway systems
Supervised learning
Fusion reactions
Textures
Processing

Cite this

Sanberg, W. P., Dubbelman, G., & With, de, P. H. N. (2014). Extending the stixel world with online self-supervised color modeling for road-versus-obstacle segmentation. In Proceedings of the 17th International IEEE Conference on Intelligent Transportation Systems (ITSC), 8-10 October 2014, Qingdao, China (pp. 1400-1407). Piscataway: Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/ITSC.2014.6957883
Sanberg, W.P. ; Dubbelman, G. ; With, de, P.H.N. / Extending the stixel world with online self-supervised color modeling for road-versus-obstacle segmentation. Proceedings of the 17th International IEEE Conference on Intelligent Transportation Systems (ITSC), 8-10 October 2014, Qingdao, China. Piscataway : Institute of Electrical and Electronics Engineers, 2014. pp. 1400-1407
@inproceedings{0c6042fd1f4c43f791ab8e4f7bc5019a,
title = "Extending the stixel world with online self-supervised color modeling for road-versus-obstacle segmentation",
abstract = "This work concentrates on vision processing for ADAS and intelligent vehicle applications. We propose a color extension to the disparity-based Stixel World method, so that the road can be robustly distinguished from obstacles with respect to erroneous disparity measurements. Our extension learns color appearance models for road and obstacle classes in an online and self-supervised fashion. The algorithm is tightly integrated within the core of the optimization process of the original Stixel World, allowing for strong fusion of the disparity and color signals. We perform an extensive evaluation, including different self-supervised learning strategies and different color models. Our newly recorded, publicly available data set is intentionally focused on challenging traffic scenes with many low-texture regions, causing numerous disparity artifacts. In this evaluation, we increase the F-score of the drivable distance from 0.86 to 0.97, compared to a tuned version of the state-of-the-art baseline method. This clearly shows that our color extension increases the robustness of the Stixel World, by reducing the number of falsely detected obstacles while not deteriorating the detection of true obstacles.",
author = "W.P. Sanberg and G. Dubbelman and {With, de}, P.H.N.",
year = "2014",
doi = "10.1109/ITSC.2014.6957883",
language = "English",
isbn = "978-1-4799-6077-4",
pages = "1400--1407",
booktitle = "Proceedings of the 17th International IEEE Conference on Intelligent Transportation Systems (ITSC), 8-10 October 2014, Qingdao, China",
publisher = "Institute of Electrical and Electronics Engineers",
address = "United States",

}

Sanberg, WP, Dubbelman, G & With, de, PHN 2014, Extending the stixel world with online self-supervised color modeling for road-versus-obstacle segmentation. in Proceedings of the 17th International IEEE Conference on Intelligent Transportation Systems (ITSC), 8-10 October 2014, Qingdao, China. Institute of Electrical and Electronics Engineers, Piscataway, pp. 1400-1407. https://doi.org/10.1109/ITSC.2014.6957883

Extending the stixel world with online self-supervised color modeling for road-versus-obstacle segmentation. / Sanberg, W.P.; Dubbelman, G.; With, de, P.H.N.

Proceedings of the 17th International IEEE Conference on Intelligent Transportation Systems (ITSC), 8-10 October 2014, Qingdao, China. Piscataway : Institute of Electrical and Electronics Engineers, 2014. p. 1400-1407.

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Extending the stixel world with online self-supervised color modeling for road-versus-obstacle segmentation

AU - Sanberg, W.P.

AU - Dubbelman, G.

AU - With, de, P.H.N.

PY - 2014

Y1 - 2014

N2 - This work concentrates on vision processing for ADAS and intelligent vehicle applications. We propose a color extension to the disparity-based Stixel World method, so that the road can be robustly distinguished from obstacles with respect to erroneous disparity measurements. Our extension learns color appearance models for road and obstacle classes in an online and self-supervised fashion. The algorithm is tightly integrated within the core of the optimization process of the original Stixel World, allowing for strong fusion of the disparity and color signals. We perform an extensive evaluation, including different self-supervised learning strategies and different color models. Our newly recorded, publicly available data set is intentionally focused on challenging traffic scenes with many low-texture regions, causing numerous disparity artifacts. In this evaluation, we increase the F-score of the drivable distance from 0.86 to 0.97, compared to a tuned version of the state-of-the-art baseline method. This clearly shows that our color extension increases the robustness of the Stixel World, by reducing the number of falsely detected obstacles while not deteriorating the detection of true obstacles.

AB - This work concentrates on vision processing for ADAS and intelligent vehicle applications. We propose a color extension to the disparity-based Stixel World method, so that the road can be robustly distinguished from obstacles with respect to erroneous disparity measurements. Our extension learns color appearance models for road and obstacle classes in an online and self-supervised fashion. The algorithm is tightly integrated within the core of the optimization process of the original Stixel World, allowing for strong fusion of the disparity and color signals. We perform an extensive evaluation, including different self-supervised learning strategies and different color models. Our newly recorded, publicly available data set is intentionally focused on challenging traffic scenes with many low-texture regions, causing numerous disparity artifacts. In this evaluation, we increase the F-score of the drivable distance from 0.86 to 0.97, compared to a tuned version of the state-of-the-art baseline method. This clearly shows that our color extension increases the robustness of the Stixel World, by reducing the number of falsely detected obstacles while not deteriorating the detection of true obstacles.

U2 - 10.1109/ITSC.2014.6957883

DO - 10.1109/ITSC.2014.6957883

M3 - Conference contribution

SN - 978-1-4799-6077-4

SP - 1400

EP - 1407

BT - Proceedings of the 17th International IEEE Conference on Intelligent Transportation Systems (ITSC), 8-10 October 2014, Qingdao, China

PB - Institute of Electrical and Electronics Engineers

CY - Piscataway

ER -

Sanberg WP, Dubbelman G, With, de PHN. Extending the stixel world with online self-supervised color modeling for road-versus-obstacle segmentation. In Proceedings of the 17th International IEEE Conference on Intelligent Transportation Systems (ITSC), 8-10 October 2014, Qingdao, China. Piscataway: Institute of Electrical and Electronics Engineers. 2014. p. 1400-1407 https://doi.org/10.1109/ITSC.2014.6957883