A steganalysis method in the DCT domain

In this paper, we propose a steganalysis method based on the principle of additive operator, which chooses non-zero AC coefficients as carriers, with secret information independent of the carrier information flow. In the proposed method, AC coefficient statistical and energy features are initially extracted and used to construct a 3D feature vector. By employing the principle of Fisher linear discriminate analysis, a flexible discriminate classifier suitable for the extracted features is designed to improve detection performance. We infer and confirm theory of change in the statistical and energetic characteristics of the AC coefficient before and after additive steganography. The effectiveness of the proposed method is proven by the experiments. Moreover, the proposed method consistently outperforms related methods.