Baden Letts Supervisor DR Jimmy Li Introduction Image forgery problem Target markets Image forgery detection techniques My project Conclusion Image from article “Spies turn to High-Tec info ops” written by Daniel Verton More advanced image manipulation programs on the market making it harder to detect image forgery It is very hard to visually see if an image has been altered More Very computer literate people simple to access image altering programs Has image forgery occurred on this picture? Image from “detection of copy move forgery in digital images” J.Friditch, D.Soukal, and J. Lucas Yes it has, this is the original image? Newspapers The legal system (government) Publishers Private use Image forgery detection techniques can be sorted into 5 different categories 1) Geometrical based techniques 2) Format based techniques 3) Camera based techniques 4) Physical based techniques 5) Pixel based techniques This is the process of using projective geometry that allow for refinement of planar surfaces. This allows for the ability to check if an image has been altered by comparing aspects of the image to known objects and or the same objects in the image. Image from “ Detecting image forgeries using geometrical cues” Lin Wu, Yang The process of detecting statistical correlations between lost information during the compression and formatting stage a digital image. Double JPEG JPEG blocking Exploits artefacts added in by specific camera characteristics Water marking Camera lens Camera sensor Camera on chip post processing Searches for anomalies in the image in relation between physical objects, light, and the camera Light Direction Light environment The process of comparing statistical anomalies provided within the image Cloning Resampling Splicing Statistical Image from “detection of copy move forgery in digital images” J.Friditch, D.Soukal, and J. Lucas Initiate a GUI (Graphical user interface) To implement multiple forgery detection technique programs This project will be continued on for future engineering students Completed a large amount of background research on image processing techniques Chosen two different image forgery techniques to implement Cloning and Splicing Began Began GUI design for the total project implementation of “cloning detection” It is always possible to get around image forgery programs as they all have scenarios that they don’t cover. The aim of this project is to cover as many scenarios as possible to make it as hard as possible for image forgery to occur Any questions Farid, H., 2008. A Survey of Image Forgery Detection. [Online] Available at: http://www.cs.albany.edu/~lsw/homepage/DIFS10_files/spm09.pdf [Accessed 12 September 2012]. Fridrich, J. S. L., 2004. Detection of Copy-Move Forgery in Digital Images. [Online] Available at: http://ws2.binghamton.edu/fridrich/Research/copymove.pdf [Accessed 14 September 2012]. Johnson, M. K. F., 2007. Detecting Photographic Composites of People. [Online] Available at: http://www.mit.edu/~kimo/publications/composite/iwdw07.pdf [Accessed 14 September 2012]. Wu, L. W., 2010. Detecting Image Forgeries using Geometric Cues. [Online] Available at: http://arxiv.org/ftp/arxiv/papers/1012/1012.3802.pdf [Accessed 12 September 2012].