Newcastle University

The Novel Image/photo Forensic Tools for fighting against child pornography project investigates the use of digital "fingerprints", which cameras and video recorders have left in photos or video clips, in the police investigations of sexual abuse of children or other cybercrime cases where image forensics are essential.

The prevalence of multimedia files (digital images  in particular) in all areas of life poses many benefits as well as challenges. Among others, the ease with which these images can be altered or obfuscated presents risk to the credibility of digital information in forensic investigations and intelligence gathering from digital devices. We are faced with questions such as:
    Given an image, can we tell which camera it was taken from?
    Is it possible to tell whether an image had been manipulated?
    Given an image and/or a camera, will it be possible to find other images available online that were taken from the same camera?
    Can we verify the authenticity of digital data through a link between digital content (e.g. digital image) and its original source (e.g. a camera)?
    Can we determine whether a child abuse image is real or created by computer?
    Can we quickly find a match of a given image against of a database of images based on “fingerprints” that are unique but may contain random errors?
    Given an image, can we extract contextual information that helps identify the perpetrators?  Can we place the digital data in the wider context of a crime scene via correlating it to other forms of evidence that are relevant, e.g. metadata of text (SMS) and image (MMS) messaging such as times and dates of delivery, contacts, instant messaging chat logs, use of wireless network access points etc.?

These questions – and many more – are encountered in increasing frequency by forensic experts all around the world. As such, there is a substantial need to develop a suite of practical tools that can be used to detect, identify and verify the source of multimedia data, as well as to ascertain whether the data has been tampered with or whether the image is computer generated.