QUT ePrints

Improved detection of centroids in aberroscope and Hartmann-Shack grid images

Iskander, D. Robert, Sulisz, Paul, Collins, Michael J., & Davis, Brett (2000) Improved detection of centroids in aberroscope and Hartmann-Shack grid images. In Association for Research in Scientific and Educator’s Meeting in Optometry, 2000.

Abstract

Purpose: An aberroscope or a Hartmann-Shack sensor is often used to determine the monochromatic aberrations of the human eye. Today, most of these instruments are equipped with a CCD video camera and a computer which calculates the wavefront error using the displacements in a grid image. These displacements are determined from the centroid locations of the grid points. Methods: The images that are acquired from the camera may be of low resolution and/or low intensity. The main factors that give rise to the insufficient quality of captured images include poor reflectance from the retina, eye aberrations, absorption and scatter within the eye, and limited light source power because of the potential for retinal light damage. Current methods for detecting the centroids in these images are essentially based on a Canny-Deriche oriented edge detecting filter, the performance of which has been found to be insufficient in some clinical applications. The limitations of current methods can be overcome by utilizing an image processing procedure that can not only reduce noise and other side-effects but also significantly improve the accuracy of the detected centroid estimates. Results: We have developed a four step technique comprised of the follower: (i) enhancement of the captured image, (ii) calculation of the initial coarse regions of interest using watershed transformation (iii) refinement of the regions using a constrained watershed, and (iv) centroid detection using shrinking procedure. We have tested this new technique on simulated images with controlled amounts of added noise in addition to real aberroscope and Hartmann-Shack images. The proposed technique performs very well and can accurately detect the centroids in the image grid for signal-to-noise ratios down to SNR-5 dB. Conclusions: Our technique can accurately determine the centroid locations in images of low resolution and intensity where Canny-Deriche filter based techniques fail.

Impact and interest:

Citation countsare sourced monthly from Scopus and Web of Science® citation databases.

These databases contain citations from different subsets of available publications and different time periods and thus the citation count from each is usually different. Some works are not in either database and no count is displayed. Scopus includes citations from articles published in 1996 onwards, and Web of Science® generally from 1980 onwards.

Citations counts from the Google Scholar™ indexing service can be viewed at the linked Google Scholar™ search.

ID Code: 12203
Item Type: Conference Item (Poster)
Additional Information: abstract published in Association for Research in Vision and Ophthalmology, 41(S), pp.427
Additional URLs:
Subjects: Australian and New Zealand Standard Research Classification > MEDICAL AND HEALTH SCIENCES (110000) > OPTOMETRY AND OPHTHALMOLOGY (111300)
Australian and New Zealand Standard Research Classification > MEDICAL AND HEALTH SCIENCES (110000) > OPTOMETRY AND OPHTHALMOLOGY (111300) > Vision Science (111303)
Australian and New Zealand Standard Research Classification > MEDICAL AND HEALTH SCIENCES (110000) > OPTOMETRY AND OPHTHALMOLOGY (111300) > Optical Technology (111302)
Divisions: Current > Research Centres > Centre for Health Research
Current > QUT Faculties and Divisions > Faculty of Health
Current > Institutes > Institute of Health and Biomedical Innovation
Copyright Owner: Copyright 2000 Association for Research in Vision and Ophthalmology
Deposited On: 25 Jan 2008
Last Modified: 10 Nov 2014 13:27

Export: EndNote | Dublin Core | BibTeX

Repository Staff Only: item control page