Exploring Feature Coupling and Model Coupling for Image Source Identification,Recently, there has been great interest in feature-based image source identification. Previous statistical learning-based methods usually regarded the identification process as a classification problem. They assumed the dependence of features and the dependence of models. However, the two assumptions are usually problematic because of the genuine coupling of features and models.To address the issues, in this paper, we propose a novel image source identification scheme.The experiments carried out on the Dresden image collection confirm the effectiveness of the proposed scheme. Via mining the feature coupling and model coupling, the identification accuracy can be significantly improved.

 

Image source identification is one of the most fundamental requirements in these scenarios, which aims to associate an image with its acquisition device.Existing source model identification approaches can be classified into three categories.

 

The apparent simple solution is the image metadata based approach, which is to investigate the EXIF header of an image.