An Improved Sensor Pattern Noise Estimation Method Based on Edge Guided Weighted Averaging

Wen Na Zhang, Yun Xia Liu, Jin Zhou, Yang Yang, Ngai Fong Law

Research output: Chapter in book / Conference proceedingConference article published in proceeding or bookAcademic researchpeer-review

Abstract

Sensor Pattern Noise (SPN) has proven to be an effective fingerprint for source camera identification. However, its estimation accuracy is still greatly affected by image contents. In this work, considering the confidence difference in varying image regions, an image edge guided weighted averaging scheme for robust SPN estimation is proposed. Firstly, the edge and non-edge regions are estimated by a Laplacian operator-based detector, based on which different weights are assigned to. Then, the improved SPN estimation is obtained by weighted averaging of image residuals. Finally, an edge guided weighted normalized cross-correlation measurement is proposed as similarity metric in source camera identification (SCI) applications. The effectiveness of the proposed method is verified by SCI experiments conducted on six models from the Dresden data set. Comparison results on different denoising algorithms and varying patch sizes reveal that performance improvement is more prominent for small image patches, which is demanding in real forensic applications.

Original languageEnglish
Title of host publicationMachine Learning for Cyber Security - Third International Conference, ML4CS 2020, Proceedings
EditorsXiaofeng Chen, Hongyang Yan, Qiben Yan, Xiangliang Zhang
PublisherSpringer Science and Business Media Deutschland GmbH
Pages405-415
Number of pages11
ISBN (Print)9783030624590
DOIs
Publication statusPublished - Nov 2020
Event3rd International Conference on Machine Learning for Cyber Security, ML4CS 2020 - Guangzhou, China
Duration: 8 Oct 202010 Oct 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12487 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference3rd International Conference on Machine Learning for Cyber Security, ML4CS 2020
CountryChina
CityGuangzhou
Period8/10/2010/10/20

Keywords

  • Edge detection
  • Sensor pattern noise
  • Source camera identification
  • Weighted averaging

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this