A measure for computing the dissimilarity for images is presented. The measure, based on information theory, considers the pixel matrices representing two images, and compares their greatest common sub-matrices. The algorithm to calculate the average area of square sub-matrices that exactly occur in both the images is described, together with its computational complexity, and an extension to accelerate its execution time is proposed. Experimental evaluation of the measure based on human perception of multiple subjects demonstrates that the measure is able to correctly discriminate (dis)similar images. Furthermore, an extensive quantitative evaluation on different kinds of image data sets shows the superiority of the measure with respect to other state-of-the-art measures in terms of retrieval precision.

A patch-based measure for image dissimilarity

Amelio A.
;
2016-01-01

Abstract

A measure for computing the dissimilarity for images is presented. The measure, based on information theory, considers the pixel matrices representing two images, and compares their greatest common sub-matrices. The algorithm to calculate the average area of square sub-matrices that exactly occur in both the images is described, together with its computational complexity, and an extension to accelerate its execution time is proposed. Experimental evaluation of the measure based on human perception of multiple subjects demonstrates that the measure is able to correctly discriminate (dis)similar images. Furthermore, an extensive quantitative evaluation on different kinds of image data sets shows the superiority of the measure with respect to other state-of-the-art measures in terms of retrieval precision.
File in questo prodotto:
File Dimensione Formato  
1-s2.0-S0925231215008905-main.pdf

Solo gestori archivio

Descrizione: Article
Tipologia: PDF editoriale
Dimensione 3.38 MB
Formato Adobe PDF
3.38 MB Adobe PDF   Visualizza/Apri   Richiedi una copia

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11564/770216
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 9
  • ???jsp.display-item.citation.isi??? 9
social impact