Share Email Print

Proceedings Paper

Quantifying image distortion based on Gabor filter bank and multiple regression analysis
Author(s): B. Ortiz-Jaramillo; J. C. Garcia-Alvarez; H. Führ; G. Castellanos-Dominguez; W. Philips
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Image quality assessment is indispensable for image-based applications. The approaches towards image quality assessment fall into two main categories: subjective and objective methods. Subjective assessment has been widely used. However, careful subjective assessments are experimentally difficult and lengthy, and the results obtained may vary depending on the test conditions. On the other hand, objective image quality assessment would not only alleviate the difficulties described above but would also help to expand the application field. Therefore, several works have been developed for quantifying the distortion presented on a image achieving goodness of fit between subjective and objective scores up to 92%. Nevertheless, current methodologies are designed assuming that the nature of the distortion is known. Generally, this is a limiting assumption for practical applications, since in a majority of cases the distortions in the image are unknown. Therefore, we believe that the current methods of image quality assessment should be adapted in order to identify and quantify the distortion of images at the same time. That combination can improve processes such as enhancement, restoration, compression, transmission, among others. We present an approach based on the power of the experimental design and the joint localization of the Gabor filters for studying the influence of the spatial/frequencies on image quality assessment. Therefore, we achieve a correct identification and quantification of the distortion affecting images. This method provides accurate scores and differentiability between distortions.

Paper Details

Date Published: 24 January 2012
PDF: 10 pages
Proc. SPIE 8293, Image Quality and System Performance IX, 82930E (24 January 2012); doi: 10.1117/12.912074
Show Author Affiliations
B. Ortiz-Jaramillo, Univ. Nacional de Colombia (Colombia)
Gent Univ. (Belgium)
J. C. Garcia-Alvarez, Univ. Nacional de Colombia (Colombia)
H. Führ, RWTH Aachen (Germany)
G. Castellanos-Dominguez, Univ. Nacional de Colombia (Colombia)
W. Philips, Gent Univ. (Belgium)

Published in SPIE Proceedings Vol. 8293:
Image Quality and System Performance IX
Frans Gaykema; Peter D. Burns, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?