Camera Model Identification Based on the Characteristic of CFA and Interpolation

Abstract

In this paper, we propose a camera-model classification method based on characteristics of color filter array (CFA) and interpolation. As CFA patterns and interpolation algorithms are different among different camera models, the artifacts introduced by CFA and interpolation can reflect model-specific to some extent. To capture the artifacts, we design a 69-D feature set and perform camera-model classification. Images from seven camera models in the Dresden Image Database are chosen as our experiment database. Experiment results show that in seven models detection, our method can do the classification with high detection accuracy from 98.39% to 99.88%.