A hybrid feature model for seam carving detection

Jingyu Ye; Yun Qing Shi

doi:10.1007/978-3-319-64185-0_7

A hybrid feature model for seam carving detection

Jingyu Ye, Yun Qing Shi

Electrical and Computer Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

4 Scopus citations

Abstract

Seam carving, as a content-aware image resizing algorithm, is widely used nowadays. In this paper, an advanced hybrid feature model is presented to reveal the trace of seam carving, especially seam carving at a low carving rate, applied to uncompressed digital images. Two groups of features are designed to capture energy variation and pixel variation caused by seam carving, respectively. As indicated by the experimental works, the state-of-the-art performance on detecting 5% and 10% carving rate cases has been improved from 81.13% and 90.26% to 85.75% and 94.87%, respectively.

Original language	English (US)
Title of host publication	Digital Forensics and Watermarking - 16th International Workshop, IWDW 2017, Proceedings
Editors	Yun-Qing Shi, Hyoung Joong Kim, Christian Kraetzer, Jana Dittmann
Publisher	Springer Verlag
Pages	77-89
Number of pages	13
ISBN (Print)	9783319641843
DOIs	https://doi.org/10.1007/978-3-319-64185-0_7
State	Published - 2017
Event	16th International Workshop on Digital Forensics and Watermarking, IWDW 2017 - Magdeburg, Germany Duration: Aug 23 2017 → Aug 25 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10431 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	16th International Workshop on Digital Forensics and Watermarking, IWDW 2017
Country/Territory	Germany
City	Magdeburg
Period	8/23/17 → 8/25/17

All Science Journal Classification (ASJC) codes

Theoretical Computer Science
General Computer Science

Keywords

Image forensics
Local derivative pattern
Markov transition probability
Seam carving detection
Support vector machine

Access to Document

10.1007/978-3-319-64185-0_7

Cite this

Ye, J., & Shi, Y. Q. (2017). A hybrid feature model for seam carving detection. In Y.-Q. Shi, H. J. Kim, C. Kraetzer, & J. Dittmann (Eds.), Digital Forensics and Watermarking - 16th International Workshop, IWDW 2017, Proceedings (pp. 77-89). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10431 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-64185-0_7

Ye, Jingyu ; Shi, Yun Qing. / A hybrid feature model for seam carving detection. Digital Forensics and Watermarking - 16th International Workshop, IWDW 2017, Proceedings. editor / Yun-Qing Shi ; Hyoung Joong Kim ; Christian Kraetzer ; Jana Dittmann. Springer Verlag, 2017. pp. 77-89 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{062f4d9ff16e4d41ac5232e99092913b,

title = "A hybrid feature model for seam carving detection",

abstract = "Seam carving, as a content-aware image resizing algorithm, is widely used nowadays. In this paper, an advanced hybrid feature model is presented to reveal the trace of seam carving, especially seam carving at a low carving rate, applied to uncompressed digital images. Two groups of features are designed to capture energy variation and pixel variation caused by seam carving, respectively. As indicated by the experimental works, the state-of-the-art performance on detecting 5% and 10% carving rate cases has been improved from 81.13% and 90.26% to 85.75% and 94.87%, respectively.",

keywords = "Image forensics, Local derivative pattern, Markov transition probability, Seam carving detection, Support vector machine",

author = "Jingyu Ye and Shi, {Yun Qing}",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2017.; 16th International Workshop on Digital Forensics and Watermarking, IWDW 2017 ; Conference date: 23-08-2017 Through 25-08-2017",

year = "2017",

doi = "10.1007/978-3-319-64185-0_7",

language = "English (US)",

isbn = "9783319641843",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "77--89",

editor = "Yun-Qing Shi and Kim, {Hyoung Joong} and Christian Kraetzer and Jana Dittmann",

booktitle = "Digital Forensics and Watermarking - 16th International Workshop, IWDW 2017, Proceedings",

address = "Germany",

}

Ye, J & Shi, YQ 2017, A hybrid feature model for seam carving detection. in Y-Q Shi, HJ Kim, C Kraetzer & J Dittmann (eds), Digital Forensics and Watermarking - 16th International Workshop, IWDW 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10431 LNCS, Springer Verlag, pp. 77-89, 16th International Workshop on Digital Forensics and Watermarking, IWDW 2017, Magdeburg, Germany, 8/23/17. https://doi.org/10.1007/978-3-319-64185-0_7

A hybrid feature model for seam carving detection. / Ye, Jingyu; Shi, Yun Qing.
Digital Forensics and Watermarking - 16th International Workshop, IWDW 2017, Proceedings. ed. / Yun-Qing Shi; Hyoung Joong Kim; Christian Kraetzer; Jana Dittmann. Springer Verlag, 2017. p. 77-89 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10431 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - A hybrid feature model for seam carving detection

AU - Ye, Jingyu

AU - Shi, Yun Qing

N1 - Publisher Copyright: © Springer International Publishing AG 2017.

PY - 2017

Y1 - 2017

N2 - Seam carving, as a content-aware image resizing algorithm, is widely used nowadays. In this paper, an advanced hybrid feature model is presented to reveal the trace of seam carving, especially seam carving at a low carving rate, applied to uncompressed digital images. Two groups of features are designed to capture energy variation and pixel variation caused by seam carving, respectively. As indicated by the experimental works, the state-of-the-art performance on detecting 5% and 10% carving rate cases has been improved from 81.13% and 90.26% to 85.75% and 94.87%, respectively.

AB - Seam carving, as a content-aware image resizing algorithm, is widely used nowadays. In this paper, an advanced hybrid feature model is presented to reveal the trace of seam carving, especially seam carving at a low carving rate, applied to uncompressed digital images. Two groups of features are designed to capture energy variation and pixel variation caused by seam carving, respectively. As indicated by the experimental works, the state-of-the-art performance on detecting 5% and 10% carving rate cases has been improved from 81.13% and 90.26% to 85.75% and 94.87%, respectively.

KW - Image forensics

KW - Local derivative pattern

KW - Markov transition probability

KW - Seam carving detection

KW - Support vector machine

UR - http://www.scopus.com/inward/record.url?scp=85028464730&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85028464730&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-64185-0_7

DO - 10.1007/978-3-319-64185-0_7

M3 - Conference contribution

AN - SCOPUS:85028464730

SN - 9783319641843

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 77

EP - 89

BT - Digital Forensics and Watermarking - 16th International Workshop, IWDW 2017, Proceedings

A2 - Shi, Yun-Qing

A2 - Kim, Hyoung Joong

A2 - Kraetzer, Christian

A2 - Dittmann, Jana

PB - Springer Verlag

T2 - 16th International Workshop on Digital Forensics and Watermarking, IWDW 2017

Y2 - 23 August 2017 through 25 August 2017

ER -

Ye J, Shi YQ. A hybrid feature model for seam carving detection. In Shi YQ, Kim HJ, Kraetzer C, Dittmann J, editors, Digital Forensics and Watermarking - 16th International Workshop, IWDW 2017, Proceedings. Springer Verlag. 2017. p. 77-89. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-64185-0_7

A hybrid feature model for seam carving detection

Abstract

Publication series

Other

All Science Journal Classification (ASJC) codes

Keywords

Access to Document

Other files and links

Fingerprint

Cite this