Connected Attribute Filtering Based on Contour Smoothness

Georgios Ouzounis; Erik R. Urbach; M.H.F. Wilkinson

Connected Attribute Filtering Based on Contour Smoothness

Georgios Ouzounis, Erik R. Urbach, M.H.F. Wilkinson

Intelligent Systems

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

300 Downloads (Pure)

Abstract

A new attribute measuring the contour smoothness of 2-D objects is presented in the context of morphological attribute filtering. The attribute is based on the ratio of the circularity and non-compactness, and has a maximum of 1 for a perfect circle. It decreases as the object boundary becomes irregular. Computation on hierarchical image representation structures relies on five auxiliary data members and is rapid. Contour smoothness is a suitable descriptor for detecting and discriminating man-made structures from other image features. An example is demonstrated on a very-high-resolution satellite image using connected pattern spectra and the switchboard platform.

Original language	English
Title of host publication	11th International Conference "Pattern Recognition and Image Analysis
Subtitle of host publication	New Information Technologies" (PRIA-11-2013)
Editors	Yu.I. Zhuravlev
Publisher	The Russian Academie of Science
Pages	286-289
Volume	I
ISBN (Print)	978-5-88940-130-8
Publication status	Published - 2013
Event	PRIA 2013 : 11th International Conference on Pattern Recognition and Image Analysis - Samara, Russian Federation Duration: 23-Sept-2013 → 28-Sept-2013

Conference

Conference	PRIA 2013
Country/Territory	Russian Federation
City	Samara
Period	23/09/2013 → 28/09/2013

Access to Document

14 Proceedings - Volume 1-IIFinal publisher's version, 19.6 MBLicence: Taverne

Handle.net

Persistent link

Cite this

@inproceedings{0f97ce13d0944267846914f81d4295e0,

title = "Connected Attribute Filtering Based on Contour Smoothness",

abstract = "A new attribute measuring the contour smoothness of 2-D objects is presented in the context of morphological attribute filtering. The attribute is based on the ratio of the circularity and non-compactness, and has a maximum of 1 for a perfect circle. It decreases as the object boundary becomes irregular. Computation on hierarchical image representation structures relies on five auxiliary data members and is rapid. Contour smoothness is a suitable descriptor for detecting and discriminating man-made structures from other image features. An example is demonstrated on a very-high-resolution satellite image using connected pattern spectra and the switchboard platform.",

author = "Georgios Ouzounis and Urbach, {Erik R.} and M.H.F. Wilkinson",

year = "2013",

language = "English",

isbn = "978-5-88940-130-8",

volume = "I",

pages = "286--289",

editor = "Yu.I. Zhuravlev",

booktitle = "11th International Conference {"}Pattern Recognition and Image Analysis",

publisher = "The Russian Academie of Science",

note = "PRIA 2013 : 11th International Conference on Pattern Recognition and Image Analysis ; Conference date: 23-09-2013 Through 28-09-2013",

}

Connected Attribute Filtering Based on Contour Smoothness. / Ouzounis, Georgios; Urbach, Erik R.; Wilkinson, M.H.F.
11th International Conference "Pattern Recognition and Image Analysis: New Information Technologies" (PRIA-11-2013). ed. / Yu.I. Zhuravlev. Vol. I The Russian Academie of Science, 2013. p. 286-289.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Connected Attribute Filtering Based on Contour Smoothness

AU - Ouzounis, Georgios

AU - Urbach, Erik R.

AU - Wilkinson, M.H.F.

PY - 2013

Y1 - 2013

N2 - A new attribute measuring the contour smoothness of 2-D objects is presented in the context of morphological attribute filtering. The attribute is based on the ratio of the circularity and non-compactness, and has a maximum of 1 for a perfect circle. It decreases as the object boundary becomes irregular. Computation on hierarchical image representation structures relies on five auxiliary data members and is rapid. Contour smoothness is a suitable descriptor for detecting and discriminating man-made structures from other image features. An example is demonstrated on a very-high-resolution satellite image using connected pattern spectra and the switchboard platform.

AB - A new attribute measuring the contour smoothness of 2-D objects is presented in the context of morphological attribute filtering. The attribute is based on the ratio of the circularity and non-compactness, and has a maximum of 1 for a perfect circle. It decreases as the object boundary becomes irregular. Computation on hierarchical image representation structures relies on five auxiliary data members and is rapid. Contour smoothness is a suitable descriptor for detecting and discriminating man-made structures from other image features. An example is demonstrated on a very-high-resolution satellite image using connected pattern spectra and the switchboard platform.

M3 - Conference contribution

SN - 978-5-88940-130-8

VL - I

SP - 286

EP - 289

BT - 11th International Conference "Pattern Recognition and Image Analysis

A2 - Zhuravlev, Yu.I.

PB - The Russian Academie of Science

T2 - PRIA 2013

Y2 - 23 September 2013 through 28 September 2013

ER -

Connected Attribute Filtering Based on Contour Smoothness

Abstract

Conference

Access to Document

Handle.net

Fingerprint

Cite this