1-D DCT Domain Analysis for JPEG Double Compression Detection

The authenticity of an evidence is a necessary and crucial requirement in forensic investigations and trials, expecially when it comes to digital evidences. In order to demonstrate the authenticity of a (digital) image, different techniques have been proposed in the last years which take advantage of some kinds of detectable anomalies on double or multiple compressed JPEG images. In this article, we present an in-depth analysis in the DCT Domain of an image exploiting the "first-digits" statistics after 1D-DCT transformation. The insights thus obtained were used to build a simple and explainable classifier able to detect doubly compressed JPEG images; this approach seems to outperform much more complex techniques in the state-of-the-art.