Road edge recognition using the Stripe Hough Transform from millimeter-wave radar images

Kun Yi Guo, Edward G. Hoare, Donya Jasteh, Xin Qing Sheng, Marina Gashinova

Research output: Contribution to journalArticlepeer-review

25 Citations (Scopus)


Millimeter-wave (MMW) radar, which is used for road feature recognition, has performance that is superior to optical cameras in terms of robustness in different weather and lighting conditions, as well as providing ranging capabilities. However, the signatures of road features in MMW radar images are quite different from that of optical images, and even physically continuous features, such as road edges, will be presented as a set of bright points or spots distributed along the roadside. Therefore, discrimination of the radar features is of paramount importance in automotive imaging systems. To tackle this problem, an approach called the stripe Hough transform (HT) is introduced in this paper, allowing enhanced extraction of the geometry of the road path. The performance of the approach is demonstrated by comparison of extracted features from MMW images with the real geometry of the road and with the results of processing by classical HT.

Original languageEnglish
Article number6882781
Pages (from-to)825-833
Number of pages9
JournalIEEE Transactions on Intelligent Transportation Systems
Issue number2
Early online date28 Aug 2014
Publication statusPublished - 1 Apr 2015


  • Adaptive cruise control (ACC)
  • Hough transform (HT)
  • millimeter-wave (MMW) radars
  • road features

ASJC Scopus subject areas

  • Automotive Engineering
  • Computer Science Applications
  • Mechanical Engineering


Dive into the research topics of 'Road edge recognition using the Stripe Hough Transform from millimeter-wave radar images'. Together they form a unique fingerprint.

Cite this