In image forensics, JPEG compression history estimation problem consists of three stages: JPEG compression detection, color model identification and subsampling, and estimation of quantization steps. Although, it is well known the importance of the phase spectrum information of digital signals and its application in image processing, the phase spectrum has been poorly explored for image forensics. In this paper, a novel method for detecting JPEG compression using statistical moments of the phase spectrum is presented. The experimentation results show that the detection rates achieved by the proposed strategy are competitive with other methods reported in the state-of-the-art. Index Terms—Image forensics, JPEG compression, Phase spectrum.

Published in: International Conference on Information Society (i-Society 2017)

• Date of Conference: 17-19 July 2017
• DOI: 10.2053/iSociety.2017.0018
• ISBN: 978-1-908320-80-3
• Conference Location: Dublin, Ireland