Multiple resolution search techniques for the Hough transform in high dimensional parameter spaces.

Research output: Chapter in Book/Report/Conference proceedingChapter

7 Citations (Scopus)

Abstract

The standard Hough transform could be used to solve recognition tasks in range data, if it were not for the high dimensional parameter spaces these tasks require. This paper describes two alternative methods which apply search techniques at multiple resolutions in order to find sets of parameters which fit the data best. Both methods are as robust as but much faster than the standard Hough transform. The first method, called recursive lattice search, employs a data structure similar to the quad-tree or oct-tree in order to attain its efficiency. The second method, called resolution hill climbing, finds a trail of hypothesized parameter sets, each of which fits the data at a higher resolution than the one preceding it. -from Author

Original languageEnglish (US)
Title of host publicationTechniques for 3-D machine perception
EditorsA. Rosenfeld
PublisherNorth-Holland; Machine Intelligence & Pattern Recognition, 3
Pages231-254
Number of pages24
StatePublished - 1986
Externally publishedYes

Fingerprint

transform
parameter
method

ASJC Scopus subject areas

  • Earth and Planetary Sciences(all)
  • Environmental Science(all)

Cite this

Milenkovic, V. (1986). Multiple resolution search techniques for the Hough transform in high dimensional parameter spaces. In A. Rosenfeld (Ed.), Techniques for 3-D machine perception (pp. 231-254). North-Holland; Machine Intelligence & Pattern Recognition, 3.

Multiple resolution search techniques for the Hough transform in high dimensional parameter spaces. / Milenkovic, Victor.

Techniques for 3-D machine perception. ed. / A. Rosenfeld. North-Holland; Machine Intelligence & Pattern Recognition, 3, 1986. p. 231-254.

Research output: Chapter in Book/Report/Conference proceedingChapter

Milenkovic, V 1986, Multiple resolution search techniques for the Hough transform in high dimensional parameter spaces. in A Rosenfeld (ed.), Techniques for 3-D machine perception. North-Holland; Machine Intelligence & Pattern Recognition, 3, pp. 231-254.
Milenkovic V. Multiple resolution search techniques for the Hough transform in high dimensional parameter spaces. In Rosenfeld A, editor, Techniques for 3-D machine perception. North-Holland; Machine Intelligence & Pattern Recognition, 3. 1986. p. 231-254
Milenkovic, Victor. / Multiple resolution search techniques for the Hough transform in high dimensional parameter spaces. Techniques for 3-D machine perception. editor / A. Rosenfeld. North-Holland; Machine Intelligence & Pattern Recognition, 3, 1986. pp. 231-254
@inbook{fe444f712aee4bb98a2bd26670c68a32,
title = "Multiple resolution search techniques for the Hough transform in high dimensional parameter spaces.",
abstract = "The standard Hough transform could be used to solve recognition tasks in range data, if it were not for the high dimensional parameter spaces these tasks require. This paper describes two alternative methods which apply search techniques at multiple resolutions in order to find sets of parameters which fit the data best. Both methods are as robust as but much faster than the standard Hough transform. The first method, called recursive lattice search, employs a data structure similar to the quad-tree or oct-tree in order to attain its efficiency. The second method, called resolution hill climbing, finds a trail of hypothesized parameter sets, each of which fits the data at a higher resolution than the one preceding it. -from Author",
author = "Victor Milenkovic",
year = "1986",
language = "English (US)",
pages = "231--254",
editor = "A. Rosenfeld",
booktitle = "Techniques for 3-D machine perception",
publisher = "North-Holland; Machine Intelligence & Pattern Recognition, 3",

}

TY - CHAP

T1 - Multiple resolution search techniques for the Hough transform in high dimensional parameter spaces.

AU - Milenkovic, Victor

PY - 1986

Y1 - 1986

N2 - The standard Hough transform could be used to solve recognition tasks in range data, if it were not for the high dimensional parameter spaces these tasks require. This paper describes two alternative methods which apply search techniques at multiple resolutions in order to find sets of parameters which fit the data best. Both methods are as robust as but much faster than the standard Hough transform. The first method, called recursive lattice search, employs a data structure similar to the quad-tree or oct-tree in order to attain its efficiency. The second method, called resolution hill climbing, finds a trail of hypothesized parameter sets, each of which fits the data at a higher resolution than the one preceding it. -from Author

AB - The standard Hough transform could be used to solve recognition tasks in range data, if it were not for the high dimensional parameter spaces these tasks require. This paper describes two alternative methods which apply search techniques at multiple resolutions in order to find sets of parameters which fit the data best. Both methods are as robust as but much faster than the standard Hough transform. The first method, called recursive lattice search, employs a data structure similar to the quad-tree or oct-tree in order to attain its efficiency. The second method, called resolution hill climbing, finds a trail of hypothesized parameter sets, each of which fits the data at a higher resolution than the one preceding it. -from Author

UR - http://www.scopus.com/inward/record.url?scp=0022823224&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0022823224&partnerID=8YFLogxK

M3 - Chapter

SP - 231

EP - 254

BT - Techniques for 3-D machine perception

A2 - Rosenfeld, A.

PB - North-Holland; Machine Intelligence & Pattern Recognition, 3

ER -