Detecting double H.264 compression based on analyzing prediction residual distribution

S. Chen; T. F. Sun; X. H. Jiang; P. S. He; S. L. Wang; Y. Q. Shi

doi:10.1007/978-3-319-53465-7_5

Detecting double H.264 compression based on analyzing prediction residual distribution

S. Chen, T. F. Sun, X. H. Jiang, P. S. He, S. L. Wang, Y. Q. Shi

Electrical and Computer Engineering

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

3 Scopus citations

Abstract

Detecting double video compression has become an important issue in video forensics. A novel double H.264 compression detection scheme based on Prediction Residual Distribution (PRED) analysis is proposed in the paper. The proposed scheme can be applied to detect double H.264 compression with non-aligned GOP structures. For each frame of a given video, the prediction residual is first calculated and the average value of the prediction residual in each non-overlapping 4 4 block is recorded to reduce the influence of the noise. Then the PRED feature is represented by the distribution of the average prediction residual in each frames. After that, the Jensen-Shannon Divergence (JSD) is introduced to measure the difference between the PRED features of adjacent two frames. Finally, a Periodic Analysis (PA) method is applied to the final feature sequence to detect double H.264 compression and to estimate the first GOP size. Fourteen public YUV sequences are adopted for evaluation. Experiments have demonstrated that the proposed scheme can achieve better performance than the state-of-the-art method investigated.

Original language	English (US)
Title of host publication	Digital Forensics and Watermarking - 15th International Workshop, IWDW 2016, Revised Selected Papers
Editors	Hyoung Joong Kim, Feng Liu, Fernando Perez-Gonzalez, Yun Qing Shi
Publisher	Springer Verlag
Pages	61-74
Number of pages	14
ISBN (Print)	9783319534640
DOIs	https://doi.org/10.1007/978-3-319-53465-7_5
State	Published - 2017
Event	15th International Workshop on Digital-Forensics and Watermarking, IWDW 2016 - Beijing, China Duration: Sep 17 2016 → Sep 19 2016

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10082 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	15th International Workshop on Digital-Forensics and Watermarking, IWDW 2016
Country/Territory	China
City	Beijing
Period	9/17/16 → 9/19/16

All Science Journal Classification (ASJC) codes

Theoretical Computer Science
General Computer Science

Keywords

Double compression detection
First GOP estimation
Non-aligned GOP structure
Prediction residual distribution

Access to Document

10.1007/978-3-319-53465-7_5

Cite this

Chen, S., Sun, T. F., Jiang, X. H., He, P. S., Wang, S. L., & Shi, Y. Q. (2017). Detecting double H.264 compression based on analyzing prediction residual distribution. In H. J. Kim, F. Liu, F. Perez-Gonzalez, & Y. Q. Shi (Eds.), Digital Forensics and Watermarking - 15th International Workshop, IWDW 2016, Revised Selected Papers (pp. 61-74). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10082 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-53465-7_5

Chen, S. ; Sun, T. F. ; Jiang, X. H. et al. / Detecting double H.264 compression based on analyzing prediction residual distribution. Digital Forensics and Watermarking - 15th International Workshop, IWDW 2016, Revised Selected Papers. editor / Hyoung Joong Kim ; Feng Liu ; Fernando Perez-Gonzalez ; Yun Qing Shi. Springer Verlag, 2017. pp. 61-74 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{3771f9d4684d4e06bef97fc67a220954,

title = "Detecting double H.264 compression based on analyzing prediction residual distribution",

abstract = "Detecting double video compression has become an important issue in video forensics. A novel double H.264 compression detection scheme based on Prediction Residual Distribution (PRED) analysis is proposed in the paper. The proposed scheme can be applied to detect double H.264 compression with non-aligned GOP structures. For each frame of a given video, the prediction residual is first calculated and the average value of the prediction residual in each non-overlapping 4 4 block is recorded to reduce the influence of the noise. Then the PRED feature is represented by the distribution of the average prediction residual in each frames. After that, the Jensen-Shannon Divergence (JSD) is introduced to measure the difference between the PRED features of adjacent two frames. Finally, a Periodic Analysis (PA) method is applied to the final feature sequence to detect double H.264 compression and to estimate the first GOP size. Fourteen public YUV sequences are adopted for evaluation. Experiments have demonstrated that the proposed scheme can achieve better performance than the state-of-the-art method investigated.",

keywords = "Double compression detection, First GOP estimation, Non-aligned GOP structure, Prediction residual distribution",

author = "S. Chen and Sun, {T. F.} and Jiang, {X. H.} and He, {P. S.} and Wang, {S. L.} and Shi, {Y. Q.}",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2017.; 15th International Workshop on Digital-Forensics and Watermarking, IWDW 2016 ; Conference date: 17-09-2016 Through 19-09-2016",

year = "2017",

doi = "10.1007/978-3-319-53465-7_5",

language = "English (US)",

isbn = "9783319534640",

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

publisher = "Springer Verlag",

pages = "61--74",

editor = "Kim, {Hyoung Joong} and Feng Liu and Fernando Perez-Gonzalez and Shi, {Yun Qing}",

booktitle = "Digital Forensics and Watermarking - 15th International Workshop, IWDW 2016, Revised Selected Papers",

address = "Germany",

}

Chen, S, Sun, TF, Jiang, XH, He, PS, Wang, SL & Shi, YQ 2017, Detecting double H.264 compression based on analyzing prediction residual distribution. in HJ Kim, F Liu, F Perez-Gonzalez & YQ Shi (eds), Digital Forensics and Watermarking - 15th International Workshop, IWDW 2016, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10082 LNCS, Springer Verlag, pp. 61-74, 15th International Workshop on Digital-Forensics and Watermarking, IWDW 2016, Beijing, China, 9/17/16. https://doi.org/10.1007/978-3-319-53465-7_5

Detecting double H.264 compression based on analyzing prediction residual distribution. / Chen, S.; Sun, T. F.; Jiang, X. H. et al.
Digital Forensics and Watermarking - 15th International Workshop, IWDW 2016, Revised Selected Papers. ed. / Hyoung Joong Kim; Feng Liu; Fernando Perez-Gonzalez; Yun Qing Shi. Springer Verlag, 2017. p. 61-74 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10082 LNCS).

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

TY - GEN

T1 - Detecting double H.264 compression based on analyzing prediction residual distribution

AU - Chen, S.

AU - Sun, T. F.

AU - Jiang, X. H.

AU - He, P. S.

AU - Wang, S. L.

AU - Shi, Y. Q.

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

PY - 2017

Y1 - 2017

N2 - Detecting double video compression has become an important issue in video forensics. A novel double H.264 compression detection scheme based on Prediction Residual Distribution (PRED) analysis is proposed in the paper. The proposed scheme can be applied to detect double H.264 compression with non-aligned GOP structures. For each frame of a given video, the prediction residual is first calculated and the average value of the prediction residual in each non-overlapping 4 4 block is recorded to reduce the influence of the noise. Then the PRED feature is represented by the distribution of the average prediction residual in each frames. After that, the Jensen-Shannon Divergence (JSD) is introduced to measure the difference between the PRED features of adjacent two frames. Finally, a Periodic Analysis (PA) method is applied to the final feature sequence to detect double H.264 compression and to estimate the first GOP size. Fourteen public YUV sequences are adopted for evaluation. Experiments have demonstrated that the proposed scheme can achieve better performance than the state-of-the-art method investigated.

AB - Detecting double video compression has become an important issue in video forensics. A novel double H.264 compression detection scheme based on Prediction Residual Distribution (PRED) analysis is proposed in the paper. The proposed scheme can be applied to detect double H.264 compression with non-aligned GOP structures. For each frame of a given video, the prediction residual is first calculated and the average value of the prediction residual in each non-overlapping 4 4 block is recorded to reduce the influence of the noise. Then the PRED feature is represented by the distribution of the average prediction residual in each frames. After that, the Jensen-Shannon Divergence (JSD) is introduced to measure the difference between the PRED features of adjacent two frames. Finally, a Periodic Analysis (PA) method is applied to the final feature sequence to detect double H.264 compression and to estimate the first GOP size. Fourteen public YUV sequences are adopted for evaluation. Experiments have demonstrated that the proposed scheme can achieve better performance than the state-of-the-art method investigated.

KW - Double compression detection

KW - First GOP estimation

KW - Non-aligned GOP structure

KW - Prediction residual distribution

UR - http://www.scopus.com/inward/record.url?scp=85013439724&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85013439724&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-53465-7_5

DO - 10.1007/978-3-319-53465-7_5

M3 - Conference contribution

AN - SCOPUS:85013439724

SN - 9783319534640

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

SP - 61

EP - 74

BT - Digital Forensics and Watermarking - 15th International Workshop, IWDW 2016, Revised Selected Papers

A2 - Kim, Hyoung Joong

A2 - Liu, Feng

A2 - Perez-Gonzalez, Fernando

A2 - Shi, Yun Qing

PB - Springer Verlag

T2 - 15th International Workshop on Digital-Forensics and Watermarking, IWDW 2016

Y2 - 17 September 2016 through 19 September 2016

ER -

Chen S, Sun TF, Jiang XH, He PS, Wang SL, Shi YQ. Detecting double H.264 compression based on analyzing prediction residual distribution. In Kim HJ, Liu F, Perez-Gonzalez F, Shi YQ, editors, Digital Forensics and Watermarking - 15th International Workshop, IWDW 2016, Revised Selected Papers. Springer Verlag. 2017. p. 61-74. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-53465-7_5

Detecting double H.264 compression based on analyzing prediction residual distribution

Abstract

Publication series

Other

All Science Journal Classification (ASJC) codes

Keywords

Access to Document

Other files and links

Fingerprint

Cite this