Wavelet-based euclidean distance for image quality assessment

Stéfane Paris (1)

1 QGAR - Querying Graphics through Analysis and Recognition

Image quality assessment (QA) mimics the user's ability in evaluating image distortion. The presented technique uses a wavelet analysis and a distance measure associated to it. As often noticed, the wavelet analysis is a good approximation of the human visual system (HVS). The main contribution of this paper is the definition of a wavelet-based euclidean distance which embeds both, the deep-structure of the images and the specific orientation of every subband. This new definition of distance allows for an effective QA almost parameterless.

Mots clés: distance measure; human visual system; image distortion; image quality assessment; wavelet-based Euclidean distance; approximation theory; image processing; wavelet transforms

Domaines: Multimédia [cs.MM]; Traitement des images [eess.IV]; Théorie de l'information [cs.IT]

Stéfane Paris. Wavelet-based euclidean distance for image quality assessment. International Symposium ELMAR-2010, Sep 2010, Zadar, Croatia. ⟨hal-01223691⟩