An Adaptive Technique for Accurate Feature Extraction from Regular and Irregular Image Data

SA Coleman, S Suganthan, BW Scotney

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We present a single multi-scale gradient-based feature extraction algorithm that can be directly applied to irregular or regular image data and hence can be used on both range and intensity images. We illustrate the accuracy of this approach using the Figure of Merit evaluation technique on real images, demonstrating that the application of this approach to both range and intensity images is more accurate than the equivalent approach of applying a gradient operator, such as Sobel, to an intensity image and, separately, the scan-line approximation approach to range images.
LanguageEnglish
Title of host publicationUnknown Host Publication
Pages911-919
Number of pages9
Volume5716
DOIs
Publication statusPublished - Sep 2009
Event15th International Conference on Image Analysis and Processing (ICIAP 2009) - Vietri sul Mare, Italy
Duration: 1 Sep 2009 → …

Conference

Conference15th International Conference on Image Analysis and Processing (ICIAP 2009)
Period1/09/09 → …

Fingerprint

Feature extraction

Keywords

  • Range Data
  • Gradient Operator
  • Feature Extraction

Cite this

@inproceedings{2f8aa5264ab246ce84a41ef82b5a62bc,
title = "An Adaptive Technique for Accurate Feature Extraction from Regular and Irregular Image Data",
abstract = "We present a single multi-scale gradient-based feature extraction algorithm that can be directly applied to irregular or regular image data and hence can be used on both range and intensity images. We illustrate the accuracy of this approach using the Figure of Merit evaluation technique on real images, demonstrating that the application of this approach to both range and intensity images is more accurate than the equivalent approach of applying a gradient operator, such as Sobel, to an intensity image and, separately, the scan-line approximation approach to range images.",
keywords = "Range Data, Gradient Operator, Feature Extraction",
author = "SA Coleman and S Suganthan and BW Scotney",
year = "2009",
month = "9",
doi = "10.1007/978-3-642-04146-4_97",
language = "English",
volume = "5716",
pages = "911--919",
booktitle = "Unknown Host Publication",

}

Coleman, SA, Suganthan, S & Scotney, BW 2009, An Adaptive Technique for Accurate Feature Extraction from Regular and Irregular Image Data. in Unknown Host Publication. vol. 5716, pp. 911-919, 15th International Conference on Image Analysis and Processing (ICIAP 2009), 1/09/09. https://doi.org/10.1007/978-3-642-04146-4_97

An Adaptive Technique for Accurate Feature Extraction from Regular and Irregular Image Data. / Coleman, SA; Suganthan, S; Scotney, BW.

Unknown Host Publication. Vol. 5716 2009. p. 911-919.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - An Adaptive Technique for Accurate Feature Extraction from Regular and Irregular Image Data

AU - Coleman, SA

AU - Suganthan, S

AU - Scotney, BW

PY - 2009/9

Y1 - 2009/9

N2 - We present a single multi-scale gradient-based feature extraction algorithm that can be directly applied to irregular or regular image data and hence can be used on both range and intensity images. We illustrate the accuracy of this approach using the Figure of Merit evaluation technique on real images, demonstrating that the application of this approach to both range and intensity images is more accurate than the equivalent approach of applying a gradient operator, such as Sobel, to an intensity image and, separately, the scan-line approximation approach to range images.

AB - We present a single multi-scale gradient-based feature extraction algorithm that can be directly applied to irregular or regular image data and hence can be used on both range and intensity images. We illustrate the accuracy of this approach using the Figure of Merit evaluation technique on real images, demonstrating that the application of this approach to both range and intensity images is more accurate than the equivalent approach of applying a gradient operator, such as Sobel, to an intensity image and, separately, the scan-line approximation approach to range images.

KW - Range Data

KW - Gradient Operator

KW - Feature Extraction

UR - http://dx.doi.org/10.1007/978-3-642-04146-4

UR - http://dx.doi.org/10.1007/978-3-642-04146-4

U2 - 10.1007/978-3-642-04146-4_97

DO - 10.1007/978-3-642-04146-4_97

M3 - Conference contribution

VL - 5716

SP - 911

EP - 919

BT - Unknown Host Publication

ER -