Feature Extraction on Range Images – A New Approach

SA Coleman, BW Scotney, S Suganthan

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

5 Citations (Scopus)

Abstract

Range images can provide an almost 3-dimensional description of a scene. Feature driven segmentation of range images has been primarily used for 3D object recognition, and hence the accuracy of the detected features is a prominent issue. Feature extraction on range images has proven to be a more complex problem than on intensity images due to both the irregular distribution of range image data and the nature of the features that are present in range images. Approaches to range image feature extraction are often scan line based approximations that carry a significant computational overhead and hence are not appropriate for real-time processing. This paper presents a design procedure for scalable first order derivative operators that can be used directly on irregularly distributed data. Hence the method is appropriate for direct use on range image data without the requirement of image preprocessing and could form the basis of algorithms of real-time robotic applications.
LanguageEnglish
Title of host publicationUnknown Host Publication
Pages1098-1103
Number of pages6
DOIs
Publication statusPublished - Apr 2007
EventIEEE International Conference on Robotics and Automation - Rome
Duration: 1 Apr 2007 → …

Conference

ConferenceIEEE International Conference on Robotics and Automation
Period1/04/07 → …

Fingerprint

Feature extraction
Object recognition
Robotics
Derivatives
Processing

Keywords

  • range images
  • feature extraction

Cite this

Coleman, SA ; Scotney, BW ; Suganthan, S. / Feature Extraction on Range Images – A New Approach. Unknown Host Publication. 2007. pp. 1098-1103
@inproceedings{a23c53e7995543f1ae88b356e49a5c83,
title = "Feature Extraction on Range Images – A New Approach",
abstract = "Range images can provide an almost 3-dimensional description of a scene. Feature driven segmentation of range images has been primarily used for 3D object recognition, and hence the accuracy of the detected features is a prominent issue. Feature extraction on range images has proven to be a more complex problem than on intensity images due to both the irregular distribution of range image data and the nature of the features that are present in range images. Approaches to range image feature extraction are often scan line based approximations that carry a significant computational overhead and hence are not appropriate for real-time processing. This paper presents a design procedure for scalable first order derivative operators that can be used directly on irregularly distributed data. Hence the method is appropriate for direct use on range image data without the requirement of image preprocessing and could form the basis of algorithms of real-time robotic applications.",
keywords = "range images, feature extraction",
author = "SA Coleman and BW Scotney and S Suganthan",
year = "2007",
month = "4",
doi = "10.1109/ROBOT.2007.363131",
language = "English",
isbn = "1-4244-0602-1",
pages = "1098--1103",
booktitle = "Unknown Host Publication",

}

Coleman, SA, Scotney, BW & Suganthan, S 2007, Feature Extraction on Range Images – A New Approach. in Unknown Host Publication. pp. 1098-1103, IEEE International Conference on Robotics and Automation, 1/04/07. https://doi.org/10.1109/ROBOT.2007.363131

Feature Extraction on Range Images – A New Approach. / Coleman, SA; Scotney, BW; Suganthan, S.

Unknown Host Publication. 2007. p. 1098-1103.

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

TY - GEN

T1 - Feature Extraction on Range Images – A New Approach

AU - Coleman, SA

AU - Scotney, BW

AU - Suganthan, S

PY - 2007/4

Y1 - 2007/4

N2 - Range images can provide an almost 3-dimensional description of a scene. Feature driven segmentation of range images has been primarily used for 3D object recognition, and hence the accuracy of the detected features is a prominent issue. Feature extraction on range images has proven to be a more complex problem than on intensity images due to both the irregular distribution of range image data and the nature of the features that are present in range images. Approaches to range image feature extraction are often scan line based approximations that carry a significant computational overhead and hence are not appropriate for real-time processing. This paper presents a design procedure for scalable first order derivative operators that can be used directly on irregularly distributed data. Hence the method is appropriate for direct use on range image data without the requirement of image preprocessing and could form the basis of algorithms of real-time robotic applications.

AB - Range images can provide an almost 3-dimensional description of a scene. Feature driven segmentation of range images has been primarily used for 3D object recognition, and hence the accuracy of the detected features is a prominent issue. Feature extraction on range images has proven to be a more complex problem than on intensity images due to both the irregular distribution of range image data and the nature of the features that are present in range images. Approaches to range image feature extraction are often scan line based approximations that carry a significant computational overhead and hence are not appropriate for real-time processing. This paper presents a design procedure for scalable first order derivative operators that can be used directly on irregularly distributed data. Hence the method is appropriate for direct use on range image data without the requirement of image preprocessing and could form the basis of algorithms of real-time robotic applications.

KW - range images

KW - feature extraction

U2 - 10.1109/ROBOT.2007.363131

DO - 10.1109/ROBOT.2007.363131

M3 - Conference contribution

SN - 1-4244-0602-1

SP - 1098

EP - 1103

BT - Unknown Host Publication

ER -