Boundary location from range data using Hough transform

Chong Sze TONG*, Y. W. Yeung

*Corresponding author for this work

Research output: Contribution to journalJournal articlepeer-review

Abstract

Sonar offers a convenient and economic means of obtaining range data. However, noise in such range data can seriously affect the detection of nearby boundaries, using the conventional circular back projection method. In this paper, we present a new formulation of the problem whereby the boundary to be detected is identified with the envelop curve of the ellipses defined by the range data. Hough transform is then applied to extract the local tangents to the boundary. The method was tested on both simulated and real data from polygonal boundaries with very promising results.

Original languageEnglish
Pages (from-to)1975-1982
Number of pages8
JournalPattern Recognition
Volume34
Issue number10
DOIs
Publication statusPublished - Oct 2001

Scopus Subject Areas

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

User-Defined Keywords

  • Boundary location
  • Circular back projection
  • Data fusion
  • Hough transform
  • Range data
  • Sonar

Fingerprint

Dive into the research topics of 'Boundary location from range data using Hough transform'. Together they form a unique fingerprint.

Cite this