This paper describes a new system for image quality assessment related to subjective visual perception. The novelties of the contribution consist in the realization of a Web-based collaborative platform, which (a) allows a set of human observers to cooperate to the process of Mean Opinion Score generation in a complete asynchronous and distance-online participation, and (b) is fully responsive, thus the images can be tested on different devices, not only monitors or television sets, but also smartphones. A new method for the analysis of the results is also proposed; it is based on visual tables, which arrange the semantic of the experimental data in colored patterns, useful for understanding the properties of both images and observers. A set of target images has been identified and included in a Web based prototype to which testers can freely connect. The system is fully customizable because the code is freely available for download; in this way the programmer may easily change all the main parameters (images, number of impairments levels, type of impairment, and so on). Experimental results are reported and commented, with considerations on the possible applications of the system in different contexts.

A Collaborative Responsive and Fully Customizable System for Image Quality Assessment based on Subjective Visual Perception

Maria Grazia Albanesi
2022

Abstract

This paper describes a new system for image quality assessment related to subjective visual perception. The novelties of the contribution consist in the realization of a Web-based collaborative platform, which (a) allows a set of human observers to cooperate to the process of Mean Opinion Score generation in a complete asynchronous and distance-online participation, and (b) is fully responsive, thus the images can be tested on different devices, not only monitors or television sets, but also smartphones. A new method for the analysis of the results is also proposed; it is based on visual tables, which arrange the semantic of the experimental data in colored patterns, useful for understanding the properties of both images and observers. A set of target images has been identified and included in a Web based prototype to which testers can freely connect. The system is fully customizable because the code is freely available for download; in this way the programmer may easily change all the main parameters (images, number of impairments levels, type of impairment, and so on). Experimental results are reported and commented, with considerations on the possible applications of the system in different contexts.
978-1-61208-978-2
File in questo prodotto:
Non ci sono file associati a questo prodotto.

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: http://hdl.handle.net/11571/1454963
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact