Please use this identifier to cite or link to this item:
Title: Extracting PRNu noise from H.264 coded videos
Authors: Altınışık, Enes
Tasdemir, Kasım
Sencar, Hüsrev Taha
Keywords: Algorithms
Image processing
Camera identification
Issue Date: 2018
Publisher: IEEE Computer SOC
Source: Altinisik, E., Tasdemir, K., & Sencar, H. T. (2018, September). Extracting Prnu Noise from H. 264 Coded Videos. In 2018 26th European Signal Processing Conference (EUSIPCO) (pp. 1367-1371). IEEE.
Series/Report no.: European Signal Processing Conference
Abstract: Every device equipped with a digital camera has a unique identity. This phenomenon is essentially due to a systematic noise component of an imaging sensor, known as photo-response non-uniformity (PRNU) noise. An imaging sensor inadvertently introduces this noise pattern to all media captured by that imaging sensor. The procedure for extracting PRNU noise has been well studied in the context of photographic images, however, its extension to video has so far been neglected. In this work, considering H.264 coding standard, we describe a procedure to extract sensor fingerprint from non-stabilized videos. The crux of our method is to remove a filtering procedure applied at the decoder to reduce blockiness and to use macroblocks selectively when estimating PRNU noise pattern. Results show that our method has a potential to improve matching performance significantly.
Description: 26th European Signal Processing Conference (2018 : Rome; Italy)
ISBN: 978-90-827970-1-5
ISSN: 2076-1465
Appears in Collections:Bilgisayar Mühendisliği Bölümü / Department of Computer Engineering
Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collection
WoS İndeksli Yayınlar Koleksiyonu / WoS Indexed Publications Collection

Show full item record

CORE Recommender


checked on Sep 23, 2022


checked on Feb 4, 2023

Page view(s)

checked on Dec 26, 2022

Google ScholarTM



Items in GCRIS Repository are protected by copyright, with all rights reserved, unless otherwise indicated.