Paper: | IMDSP-P13.4 | ||
Session: | Image and Video Coding and Quality | ||
Time: | Friday, May 21, 15:30 - 17:30 | ||
Presentation: | Poster | ||
Topic: | Image and Multidimensional Signal Processing: Image Quality | ||
Title: | IMAGE INFORMATION AND VISUAL QUALITY | ||
Authors: | Hamid Sheikh; University of Texas, Austin | ||
Alan Bovik; University of Texas, Austin | |||
Abstract: | Measurement of image quality is crucial for many image-processingalgorithms. Traditionally, image quality assessment algorithmspredict visual quality by comparing a distorted image against areference image, typically by modelling the Human Visual System(HVS), or by using arbitrary signal fidelity criteria. In thispaper we adopt a new paradigm for image quality assessment. Wepropose an information fidelity criterion that quantifies theShannon information that is shared between the reference and thedistorted images relative to the information contained in thereference image itself. We use Natural Scene Statistics (NSS)modelling in concert with an image degradation model and an HVSmodel. We demonstrate the performance of our algorithm by testingit on a data set of 779 images, and show that our method iscompetitive with state of the art quality assessment methods, andoutperforms them in our simulations. | ||
Back |
Home -||-
Organizing Committee -||-
Technical Committee -||-
Technical Program -||-
Plenaries
Paper Submission -||-
Special Sessions -||-
ITT -||-
Paper Review -||-
Exhibits -||-
Tutorials
Information -||-
Registration -||-
Travel Insurance -||-
Housing -||-
Workshops