An enhanced approach for detecting double JPEG compression with the same quantization matrix

作者:

Highlights:

摘要

The number of different JPEG coefficients between two successive JPEG compressed images with the same quantization matrix will monotonically decrease in general. This intrinsic property can be used to detect double JPEG compression with the same quantization matrix. However, such a decreasing tendency is not significant for some JPEG images compressed with low quality factors (QFs). It is especially difficult to detect double JPEG compression in this scenario. By analyzing the statistical properties of JPEG coefficients with various QFs, an enhanced random perturbation based double JPEG compression detection method is proposed in this paper. Unlike existing random perturbation strategy that just indiscriminately selects JPEG coefficients for modification, in the proposed method, only JPEG coefficients with values ±1 are modified by a novel modification strategy. Experimental results demonstrate that the proposed method outperforms some state-of-the-art detection methods, including the original random perturbation based one, especially for the case of low QFs.

论文关键词:Digital forensics,Double JPEG compression,JPEG coefficients,Quantization matrix

论文评审过程:Received 30 September 2018, Revised 17 April 2019, Accepted 17 April 2019, Available online 30 May 2019, Version of Record 30 May 2019.

论文官网地址:https://doi.org/10.1016/j.image.2019.04.016