Experimental study of Print-and-Scan impact as random process

Iuliia Tkachenko 1 William Puech 1 Olivier Strauss 1 Christophe Destruel
1 ICAR - Image & Interaction
LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier
Abstract : The Print-and-Scan (P&S) process is generally considered as being a random process that can be modeled by a white additive normal process, ergodic in the wide sense. This study aims at experimentally validate or invalidate this hypothesis. Moreover, the experiments we carried on have been conducted in order to separate the printing from scanning impact. The main conclusion of these experiments is that the usual hypothesis cannot be supported by the experiments. The normal modeling can be invalidated for the whole P&S process. Moreover, it has been highlighted that this process cannot be considered as being neither white nor ergodic in the wide sense. The scanner noise seems to be mean ergodic and closer to a Laplace, than to a normal distribution.
Type de document :
Communication dans un congrès
Electronic Imaging , Feb 2016, San Francisco, United States. 2016, Media Watermarking, Security, and Forensics. 〈http://www.imaging.org/ist/conferences/ei2016/C-MWSF.cfm〉. 〈10.2352/ISSN.2470-1173.2016.8.MWSF-089〉
Liste complète des métadonnées

https://hal-lirmm.ccsd.cnrs.fr/lirmm-01379547
Contributeur : William Puech <>
Soumis le : mardi 11 octobre 2016 - 16:47:16
Dernière modification le : mardi 16 octobre 2018 - 07:52:01

Identifiants

Collections

Citation

Iuliia Tkachenko, William Puech, Olivier Strauss, Christophe Destruel. Experimental study of Print-and-Scan impact as random process. Electronic Imaging , Feb 2016, San Francisco, United States. 2016, Media Watermarking, Security, and Forensics. 〈http://www.imaging.org/ist/conferences/ei2016/C-MWSF.cfm〉. 〈10.2352/ISSN.2470-1173.2016.8.MWSF-089〉. 〈lirmm-01379547〉

Partager

Métriques

Consultations de la notice

80