Wet area and puddle detection for Advanced Driver Assistance Systems (ADAS) using a stereo camera

Jisu Kim, Jeonghyun Baek, Hyukdoo Choi, Euntai Kim

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

Wet area or puddle detection is one of the key issues for safe driving and future Advanced Driver Assistance Systems (ADAS). A new methodology for the detection of wet areas and puddles using a stereo camera is presented in this paper. Because wet areas and puddles have different characteristics, the two areas are separately treated and different detection algorithms are proposed. For the detection of wet areas, color information is used for hypothesis generation (HG) and a support vector machine (SVM) is employed for hypothesis verification (HV). In HV, three features are proposed for classification; these are the polarization difference, graininess and gradient magnitude. For the detection of puddles, the depth map obtained by a stereo camera is used to exploit the fact that abrupt depth changes are detected around the puddles. In the experiment, it is shown that the proposed methods have a robust performance for the detection of wet areas or puddles.

Original languageEnglish
Pages (from-to)263-271
Number of pages9
JournalInternational Journal of Control, Automation and Systems
Volume14
Issue number1
DOIs
Publication statusPublished - 2016 Feb 1

Fingerprint

Advanced driver assistance systems
Cameras
Support vector machines
Polarization
Color
Experiments

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Computer Science Applications

Cite this

@article{cf925e969e804c7b955af13b03585f1c,
title = "Wet area and puddle detection for Advanced Driver Assistance Systems (ADAS) using a stereo camera",
abstract = "Wet area or puddle detection is one of the key issues for safe driving and future Advanced Driver Assistance Systems (ADAS). A new methodology for the detection of wet areas and puddles using a stereo camera is presented in this paper. Because wet areas and puddles have different characteristics, the two areas are separately treated and different detection algorithms are proposed. For the detection of wet areas, color information is used for hypothesis generation (HG) and a support vector machine (SVM) is employed for hypothesis verification (HV). In HV, three features are proposed for classification; these are the polarization difference, graininess and gradient magnitude. For the detection of puddles, the depth map obtained by a stereo camera is used to exploit the fact that abrupt depth changes are detected around the puddles. In the experiment, it is shown that the proposed methods have a robust performance for the detection of wet areas or puddles.",
author = "Jisu Kim and Jeonghyun Baek and Hyukdoo Choi and Euntai Kim",
year = "2016",
month = "2",
day = "1",
doi = "10.1007/s12555-015-0024-0",
language = "English",
volume = "14",
pages = "263--271",
journal = "International Journal of Control, Automation and Systems",
issn = "1598-6446",
publisher = "Institute of Control, Robotics and Systems",
number = "1",

}

Wet area and puddle detection for Advanced Driver Assistance Systems (ADAS) using a stereo camera. / Kim, Jisu; Baek, Jeonghyun; Choi, Hyukdoo; Kim, Euntai.

In: International Journal of Control, Automation and Systems, Vol. 14, No. 1, 01.02.2016, p. 263-271.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Wet area and puddle detection for Advanced Driver Assistance Systems (ADAS) using a stereo camera

AU - Kim, Jisu

AU - Baek, Jeonghyun

AU - Choi, Hyukdoo

AU - Kim, Euntai

PY - 2016/2/1

Y1 - 2016/2/1

N2 - Wet area or puddle detection is one of the key issues for safe driving and future Advanced Driver Assistance Systems (ADAS). A new methodology for the detection of wet areas and puddles using a stereo camera is presented in this paper. Because wet areas and puddles have different characteristics, the two areas are separately treated and different detection algorithms are proposed. For the detection of wet areas, color information is used for hypothesis generation (HG) and a support vector machine (SVM) is employed for hypothesis verification (HV). In HV, three features are proposed for classification; these are the polarization difference, graininess and gradient magnitude. For the detection of puddles, the depth map obtained by a stereo camera is used to exploit the fact that abrupt depth changes are detected around the puddles. In the experiment, it is shown that the proposed methods have a robust performance for the detection of wet areas or puddles.

AB - Wet area or puddle detection is one of the key issues for safe driving and future Advanced Driver Assistance Systems (ADAS). A new methodology for the detection of wet areas and puddles using a stereo camera is presented in this paper. Because wet areas and puddles have different characteristics, the two areas are separately treated and different detection algorithms are proposed. For the detection of wet areas, color information is used for hypothesis generation (HG) and a support vector machine (SVM) is employed for hypothesis verification (HV). In HV, three features are proposed for classification; these are the polarization difference, graininess and gradient magnitude. For the detection of puddles, the depth map obtained by a stereo camera is used to exploit the fact that abrupt depth changes are detected around the puddles. In the experiment, it is shown that the proposed methods have a robust performance for the detection of wet areas or puddles.

UR - http://www.scopus.com/inward/record.url?scp=84957976700&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84957976700&partnerID=8YFLogxK

U2 - 10.1007/s12555-015-0024-0

DO - 10.1007/s12555-015-0024-0

M3 - Article

AN - SCOPUS:84957976700

VL - 14

SP - 263

EP - 271

JO - International Journal of Control, Automation and Systems

JF - International Journal of Control, Automation and Systems

SN - 1598-6446

IS - 1

ER -