Inference of a compact representation of sensor fingerprint for source camera identification

Ruizhe Li, Chang Tsun Li, Yu Guan

Research output: Contribution to journalArticle

  • 2 Citations

Abstract

Sensor pattern noise (SPN) is an inherent fingerprint of imaging devices, which provides an effective way for source camera identification (SCI). Although SPNs extracted from large image blocks usually yield high identification accuracy, their high dimensionality would incur a high computational cost in the matching stage, consequently hindering many applications that require efficient camera matchings. In this work, we employ and evaluate the concept of principal component analysis (PCA) de-noising in SCI tasks. Based on this concept, we present a framework that formulates a compact SPN representation. To enhance the de-noising effect, we introduce a training set construction procedure that minimizes the impact of various interfering artifacts, which is especially useful in some challenging cases, e.g., when only textured reference images are available. To further boost the SCI performance, a novel approach based on linear discriminant analysis (LDA) is adopted to extract more discriminant SPN features. To evaluate our methods, extensive experiments are conducted on the Dresden image database. The results indicate that the proposed framework can serve as an effective post-processing procedure, which not only boosts the performance, but also greatly reduces the computational cost in the matching phase.

LanguageEnglish
Pages556-567
Number of pages12
JournalPattern Recognition
Volume74
Early online dateSep 2017
DOIs
StatePublished - 01 Feb 2018

Fingerprint

Cameras
Sensors
Phase matching
Discriminant analysis
Principal component analysis
Costs
Imaging techniques
Processing
Experiments

Cite this

@article{997ba5c5c76a462fa66f71110383519a,
title = "Inference of a compact representation of sensor fingerprint for source camera identification",
abstract = "Sensor pattern noise (SPN) is an inherent fingerprint of imaging devices, which provides an effective way for source camera identification (SCI). Although SPNs extracted from large image blocks usually yield high identification accuracy, their high dimensionality would incur a high computational cost in the matching stage, consequently hindering many applications that require efficient camera matchings. In this work, we employ and evaluate the concept of principal component analysis (PCA) de-noising in SCI tasks. Based on this concept, we present a framework that formulates a compact SPN representation. To enhance the de-noising effect, we introduce a training set construction procedure that minimizes the impact of various interfering artifacts, which is especially useful in some challenging cases, e.g., when only textured reference images are available. To further boost the SCI performance, a novel approach based on linear discriminant analysis (LDA) is adopted to extract more discriminant SPN features. To evaluate our methods, extensive experiments are conducted on the Dresden image database. The results indicate that the proposed framework can serve as an effective post-processing procedure, which not only boosts the performance, but also greatly reduces the computational cost in the matching phase.",
keywords = "Image forensics, PCA de-noising, Sensor pattern noise (SPN), Source camera identification (SCI)",
author = "Ruizhe Li and Li, {Chang Tsun} and Yu Guan",
year = "2018",
month = "2",
day = "1",
doi = "10.1016/j.patcog.2017.09.027",
language = "English",
volume = "74",
pages = "556--567",
journal = "Pattern Recognition",
issn = "0031-3203",
publisher = "Elsevier",

}

Inference of a compact representation of sensor fingerprint for source camera identification. / Li, Ruizhe; Li, Chang Tsun; Guan, Yu.

In: Pattern Recognition, Vol. 74, 01.02.2018, p. 556-567.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Inference of a compact representation of sensor fingerprint for source camera identification

AU - Li,Ruizhe

AU - Li,Chang Tsun

AU - Guan,Yu

PY - 2018/2/1

Y1 - 2018/2/1

N2 - Sensor pattern noise (SPN) is an inherent fingerprint of imaging devices, which provides an effective way for source camera identification (SCI). Although SPNs extracted from large image blocks usually yield high identification accuracy, their high dimensionality would incur a high computational cost in the matching stage, consequently hindering many applications that require efficient camera matchings. In this work, we employ and evaluate the concept of principal component analysis (PCA) de-noising in SCI tasks. Based on this concept, we present a framework that formulates a compact SPN representation. To enhance the de-noising effect, we introduce a training set construction procedure that minimizes the impact of various interfering artifacts, which is especially useful in some challenging cases, e.g., when only textured reference images are available. To further boost the SCI performance, a novel approach based on linear discriminant analysis (LDA) is adopted to extract more discriminant SPN features. To evaluate our methods, extensive experiments are conducted on the Dresden image database. The results indicate that the proposed framework can serve as an effective post-processing procedure, which not only boosts the performance, but also greatly reduces the computational cost in the matching phase.

AB - Sensor pattern noise (SPN) is an inherent fingerprint of imaging devices, which provides an effective way for source camera identification (SCI). Although SPNs extracted from large image blocks usually yield high identification accuracy, their high dimensionality would incur a high computational cost in the matching stage, consequently hindering many applications that require efficient camera matchings. In this work, we employ and evaluate the concept of principal component analysis (PCA) de-noising in SCI tasks. Based on this concept, we present a framework that formulates a compact SPN representation. To enhance the de-noising effect, we introduce a training set construction procedure that minimizes the impact of various interfering artifacts, which is especially useful in some challenging cases, e.g., when only textured reference images are available. To further boost the SCI performance, a novel approach based on linear discriminant analysis (LDA) is adopted to extract more discriminant SPN features. To evaluate our methods, extensive experiments are conducted on the Dresden image database. The results indicate that the proposed framework can serve as an effective post-processing procedure, which not only boosts the performance, but also greatly reduces the computational cost in the matching phase.

KW - Image forensics

KW - PCA de-noising

KW - Sensor pattern noise (SPN)

KW - Source camera identification (SCI)

UR - http://www.scopus.com/inward/record.url?scp=85032262573&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85032262573&partnerID=8YFLogxK

U2 - 10.1016/j.patcog.2017.09.027

DO - 10.1016/j.patcog.2017.09.027

M3 - Article

VL - 74

SP - 556

EP - 567

JO - Pattern Recognition

T2 - Pattern Recognition

JF - Pattern Recognition

SN - 0031-3203

ER -