A Correlation Investigation of Anti-Aliasing Methods and Edge Detection

Classement des Méthodes de Décrénelage à l’Aide de Detection des Bords

PDF

Authors
  1. Engels, J.E.
  2. Niall, K.K.
  3. Obidowski, R.M.
Corporate Authors
Defence R&D Canada - Toronto, Toronto ONT (CAN);Array Systems Computing Inc, Toronto Ont (CAN)
Abstract
Operations that blur edges, specifically anti-aliasing methods, can improve the fidelity of digital images. We propose a technique to rank anti-aliasing methods using edge detection methods and correlation coefficients. Images were gathered for a test set and a larger second test set of images as part of a test/retest methodology. Anti-aliasing methods followed by edge detection methods were applied on both test sets and correlations were calculated between the edge-detected aliased and anti-aliased images using white pixel counts. This technique works for ranking anti-aliasing methods and is robust. It provides a consistent ranking over two test sets and it can be used to determine which anti-aliasing method offers the best blending of pixels without the loss of fidelity or content in the images.

Il y a un résumé en français ici.

Keywords
Anti-aliasing;edge detection;image processing
Report Number
DRDC-TORONTO-TR-2007-142 — Technical Report
Date of publication
03 Dec 2007
Number of Pages
112
DSTKIM No
CA030369
CANDIS No
528874
Format(s):
Electronic Document(PDF)

Permanent link

Document 1 of 1

Date modified: