Research project

Image Forensics

Image Forensics is a rather young field of research that plays an important role in the field of digital forensics due to today's fast growing availability of digital footage. Digital image data can be tampered in an easy and low-cost fashion. Relevant objects may be removed from or added to the depicted scene, and thus new photographic evidence can be produced. This creates a need for expert assessment that indisputably distinguishes forgeries and original images. Computer-assisted methods are able to obtain relevant information e.g. from distinctive traces on the data level or based on inconsistencies in the depicted scene or the image formation process.

In this research project, the current state of the art is to be advanced in several ways. Currently available methods are evaluated and improved based on real forgery data. For this, an elaborate image manipulation dataset is created and maintained. Furthermore, new methods are to be developed that concentrate on the physical consistency of the depicted scene. Properties like illumination and cast shadows are to be estimated in a robust fashion to assess the consistency of the depicted scene.

Project manager:
Dr.-Ing. Christian Riess

Project participants:
Dr.-Ing. Christian Riess

Keywords:
Image Forgery; Digital Forensics; Computer Vision

Start: 1.5.2009

Contact:
Riess, Christian
Phone +49 9131 85 27774, Fax +49 9131 85 27270, E-Mail: christian.riess@fau.de

Publications

Christlein, Vincent ; Riess, Christian ; Angelopoulou, Elli: A Study on Features for the Detection of Copy-Move Forgeries. In: Freiling, Felix (Ed.) : Sicherheit 2010 - Sicherheit, Schutz und Zuverlässigkeit (Sicherheit 2010 Berlin 5.10.-7.10.2010). 1. Edition Heidelberg : Springer, 2010, pp 105-116. (Lecture Notes in Computer Science (LNCS) Vol. P-170)

Christlein, Vincent ; Riess, Christian ; Angelopoulou, Elli: On Rotation Invariance in Copy-Move Forgery Detection. In: IEEE (Org.) : Proceedings of the 2010 Second IEEE Workshop on (Workshop on Information Forensics and Security Seattle, USA 12.12.-15.12.2010). 2010, pp -.

Riess, Christian ; Angelopoulou, Elli: Scene Illumination as an Indicator of Image Manipulation. In: Boehme, Rainer ; Fong, Philipp ; Safavi-Naini, Rei (Ed.) : Information Hiding, 6th International Workshop (Information Hiding, 6th International Workshop Calgary, Canada 28.6.-30.6.2010). Heidelberg : Springer, 2010, pp 66-80. (Lecture Notes in Computer Science (LNCS) Vol. 6387)

Christlein, Vincent ; Riess, Christian ; Jordan, Johannes ; Riess, Corinna ; Angelopoulou, Elli: An Evaluation of Popular Copy-Move Forgery Detection Approaches. In: IEEE Transactions on Information Forensics and Security 7 (2012), No. 6, pp 1841-1854
[doi>10.1109/TIFS.2012.2218597]

Zach, Fabian ; Riess, Christian ; Angelopoulou, Elli: Automated Image Forgery Detection through Classification of JPEG Ghosts. In: Axel Pinz ; Thomas Pock ; Horst Bischof ; Franz Leberl (Ed.) : Pattern Recognition (Joint 34th DAGM and 36th OAGM Symposium Graz, Austria 28.08.2012). Berlin, Heidelberg : Springer, 2012, pp 185-194. (Lecture Notes in Computer Science Vol. 7476)
[doi>10.1007/978-3-642-32717-9_19]

Riess, Christian: Einführung in die Multimediaforensik. In: Dewald, Andreas ; Freiling, Felix (Ed.) : Forensische Informatik. 2. Edition Norderstedt : BoD - Books on Demand, 2015, pp 117-150. - ISBN 9783842379473

Riess, Christian ; Pfaller, Sven ; Angelopoulou, Elli: Reflectance Normalization in Illumination-Based Image Manipulation Detection. In: Santana, Modesto Castrillon ; Kirchner, Matthias ; Riccio, Daniel ; Verdoliva, Luisa (Ed.) : BioFor 2015 - International Workshop on Recent Advances in Digital Security: Biometrics and Forensics (BioFor 2015 - International Workshop on Recent Advances in Digital Security: Biometrics and Forensics Genova 07.09.2015). Berlin : Springer, 2015, pp 3-10. (Lecture Notes in Computer Science Vol. 9281)
[doi>10.1007/978-3-319-23222-5_1]

Kirchner, Matthias ; Schöttele, Pascal ; Riess, Christian: Thinking Beyond the Block: Block Matching for Copy-Move Forgery Detection Revisited. In: Alattar, Adnan M. ; Memon, Nasir D. ; Heitzenrater, Chad D. (Ed.) : Proc. SPIE 9409, Media Watermarking, Security, and Forensics 2015 (Proc. SPIE 9409, Media Watermarking, Security, and Forensics 2015 San Francisco, CA, USA 9.2.-11.2.2015). 2015, pp 940903. (Proceedings of hte SPIE Vol. 9409)
[doi>10.1117/12.2082789]

de Carvalho, Tiago J. ; Riess, Christian ; Angelopoulou, Elli ; Pedrini, Helio ; Rocha, Anderson: Exposing Digital Image Forgeries by Illumination Color Classification. In: IEEE Transactions on Information Forensics and Security 8 (2013), No. 7, pp 1182-1194
[doi>10.1109/TIFS.2013.2265677]

Gloe, Thomas ; Kirchner, Matthias ; Riess, Christian: How we stopped worrying about content and love the metadata. Erlangen : Friedrich-Alexander Universität. 2013. - Research report. 4 pages (First IEEE Information Forensics and Security Technical Committee (IFS-TC) Image Forensics Challenge)

Riess, Christian: Physics-based and Statistical Features for Image Forensics. Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Ph.D. thesis, 2012. - 226 pages.