Zipf’s Law for Image Forensics
- Monday 1 November 2010, 13:30 to 14:30
- Open to:
- Staff, Students
- Mr Zhao Xi
Zipf’s law, one of the empirical laws, was originally used to analyse the probability of occurrence of an event in mathematical fashion. For instance, it can be used to describe the relationships between the popularity rank of words and their frequency of use in a natural language. Similarly, it can be shown that there is a mathematical pattern between the size of the population in a country and the size of its cities.
Digital watermarking techniques are considered to be active approaches because of their requirement to embed watermark bits into the content. In contrast, image forensics are essentially passive techniques, which have been developed to analyse images based solely on the image data itself. These techniques can be classified into six different categories: source classification; device identification; images linking to source device; processing history recovery; integrity detection; and anomaly investigation.
Processing history recovery relates to the partial recovery of the processing chain associated with an image. This area of image forensics focuses on detecting non-malicious alterations in an image such as lossy compression (JPEG, JPEG2000), resizing, and colour/contrast adjustments. In this talk, we propose a scheme that utilises Zipf’s Law to detect the unknown JPEG compression. Furthermore, we also present some experimental results investigating the strength of the Zipf’s law for authentication in image forensics