Application of W-Band FMCW Radar for Road Curvature Estimation in Poor Visibility Conditions

Tae Yun Lee, Vladimir Skvortsov, Myung Sik Kim, Seung Hoon Han, Min-Ho Ka

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

This paper discusses radar-based technologies that provide drivers with information about road curves under conditions in which optical sensors are unable to perform. The development of a task capable of estimating road curves has become a popular challenge since the commencement of the active development of advanced driver assistance systems and autonomous vehicles. Although the hardware implementation of a road curvature measurement system may have many variations, including cameras, we consider microwave sensors to be a reasonable alternative that offers many benefits. Road curvature measurement requires fast acquisition of the necessary data from a moving vehicle followed by efficient postprocessing of the data to identify a sufficient number of reliable featured points in the local environment. These points would enable a hypothesis about the dimensions and shape of the probable road geometry to be built. Different studies attempted to solve this problem using specialized equipment and custom-built algorithms. Our solution is to use a commercially available radar and some preprocessing and postprocessing algorithms for data conditioning and analysis to obtain reliable results in both day- and night-time environments and under various weather conditions in a relatively short time using a simplified approach based on circle-fitting algorithms. We confirmed the effectiveness of the proposed approach by conducting multiple road experiments with a commercially available microwave sensor. Considering that the road infrastructure is currently not optimized for road curvature measurements, the results we obtained with our proposed method (within 10%-17% of those on the actual map) are acceptable.

Original languageEnglish
Pages (from-to)5300-5312
Number of pages13
JournalIEEE Sensors Journal
Volume18
Issue number13
DOIs
Publication statusPublished - 2018 Jul 1

Fingerprint

roads
visibility
Visibility
Microwave sensors
radar
Radar
curvature
Advanced driver assistance systems
Optical sensors
microwave sensors
Cameras
Hardware
vehicles
Geometry
daytime
optical measuring instruments
curves
preprocessing
conditioning
Experiments

All Science Journal Classification (ASJC) codes

  • Instrumentation
  • Electrical and Electronic Engineering

Cite this

Lee, Tae Yun ; Skvortsov, Vladimir ; Kim, Myung Sik ; Han, Seung Hoon ; Ka, Min-Ho. / Application of W-Band FMCW Radar for Road Curvature Estimation in Poor Visibility Conditions. In: IEEE Sensors Journal. 2018 ; Vol. 18, No. 13. pp. 5300-5312.
@article{f615e7b5395c4117ae8bb8b6637af6a6,
title = "Application of W-Band FMCW Radar for Road Curvature Estimation in Poor Visibility Conditions",
abstract = "This paper discusses radar-based technologies that provide drivers with information about road curves under conditions in which optical sensors are unable to perform. The development of a task capable of estimating road curves has become a popular challenge since the commencement of the active development of advanced driver assistance systems and autonomous vehicles. Although the hardware implementation of a road curvature measurement system may have many variations, including cameras, we consider microwave sensors to be a reasonable alternative that offers many benefits. Road curvature measurement requires fast acquisition of the necessary data from a moving vehicle followed by efficient postprocessing of the data to identify a sufficient number of reliable featured points in the local environment. These points would enable a hypothesis about the dimensions and shape of the probable road geometry to be built. Different studies attempted to solve this problem using specialized equipment and custom-built algorithms. Our solution is to use a commercially available radar and some preprocessing and postprocessing algorithms for data conditioning and analysis to obtain reliable results in both day- and night-time environments and under various weather conditions in a relatively short time using a simplified approach based on circle-fitting algorithms. We confirmed the effectiveness of the proposed approach by conducting multiple road experiments with a commercially available microwave sensor. Considering that the road infrastructure is currently not optimized for road curvature measurements, the results we obtained with our proposed method (within 10{\%}-17{\%} of those on the actual map) are acceptable.",
author = "Lee, {Tae Yun} and Vladimir Skvortsov and Kim, {Myung Sik} and Han, {Seung Hoon} and Min-Ho Ka",
year = "2018",
month = "7",
day = "1",
doi = "10.1109/JSEN.2018.2837875",
language = "English",
volume = "18",
pages = "5300--5312",
journal = "IEEE Sensors Journal",
issn = "1530-437X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "13",

}

Application of W-Band FMCW Radar for Road Curvature Estimation in Poor Visibility Conditions. / Lee, Tae Yun; Skvortsov, Vladimir; Kim, Myung Sik; Han, Seung Hoon; Ka, Min-Ho.

In: IEEE Sensors Journal, Vol. 18, No. 13, 01.07.2018, p. 5300-5312.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Application of W-Band FMCW Radar for Road Curvature Estimation in Poor Visibility Conditions

AU - Lee, Tae Yun

AU - Skvortsov, Vladimir

AU - Kim, Myung Sik

AU - Han, Seung Hoon

AU - Ka, Min-Ho

PY - 2018/7/1

Y1 - 2018/7/1

N2 - This paper discusses radar-based technologies that provide drivers with information about road curves under conditions in which optical sensors are unable to perform. The development of a task capable of estimating road curves has become a popular challenge since the commencement of the active development of advanced driver assistance systems and autonomous vehicles. Although the hardware implementation of a road curvature measurement system may have many variations, including cameras, we consider microwave sensors to be a reasonable alternative that offers many benefits. Road curvature measurement requires fast acquisition of the necessary data from a moving vehicle followed by efficient postprocessing of the data to identify a sufficient number of reliable featured points in the local environment. These points would enable a hypothesis about the dimensions and shape of the probable road geometry to be built. Different studies attempted to solve this problem using specialized equipment and custom-built algorithms. Our solution is to use a commercially available radar and some preprocessing and postprocessing algorithms for data conditioning and analysis to obtain reliable results in both day- and night-time environments and under various weather conditions in a relatively short time using a simplified approach based on circle-fitting algorithms. We confirmed the effectiveness of the proposed approach by conducting multiple road experiments with a commercially available microwave sensor. Considering that the road infrastructure is currently not optimized for road curvature measurements, the results we obtained with our proposed method (within 10%-17% of those on the actual map) are acceptable.

AB - This paper discusses radar-based technologies that provide drivers with information about road curves under conditions in which optical sensors are unable to perform. The development of a task capable of estimating road curves has become a popular challenge since the commencement of the active development of advanced driver assistance systems and autonomous vehicles. Although the hardware implementation of a road curvature measurement system may have many variations, including cameras, we consider microwave sensors to be a reasonable alternative that offers many benefits. Road curvature measurement requires fast acquisition of the necessary data from a moving vehicle followed by efficient postprocessing of the data to identify a sufficient number of reliable featured points in the local environment. These points would enable a hypothesis about the dimensions and shape of the probable road geometry to be built. Different studies attempted to solve this problem using specialized equipment and custom-built algorithms. Our solution is to use a commercially available radar and some preprocessing and postprocessing algorithms for data conditioning and analysis to obtain reliable results in both day- and night-time environments and under various weather conditions in a relatively short time using a simplified approach based on circle-fitting algorithms. We confirmed the effectiveness of the proposed approach by conducting multiple road experiments with a commercially available microwave sensor. Considering that the road infrastructure is currently not optimized for road curvature measurements, the results we obtained with our proposed method (within 10%-17% of those on the actual map) are acceptable.

UR - http://www.scopus.com/inward/record.url?scp=85047014199&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85047014199&partnerID=8YFLogxK

U2 - 10.1109/JSEN.2018.2837875

DO - 10.1109/JSEN.2018.2837875

M3 - Article

AN - SCOPUS:85047014199

VL - 18

SP - 5300

EP - 5312

JO - IEEE Sensors Journal

JF - IEEE Sensors Journal

SN - 1530-437X

IS - 13

ER -