Intelligent Engineering Systems through Artificial Neural Networks, Volume 16
82 Detect JPEG Steganography Using Polynomial Fitting
Download citation file:
- Ris (Zotero)
- Reference Manager
In this paper, we present a scheme of steganalysis of JPEG images with the use of polynomial fitting. Based on the Generalized Gaussian Distribution (GGD) model in the quantized DCT coefficients, the errors between the logarithmic domain of the histogram of the DCT coefficients and the polynomial fitting are extracted as features to discriminate the adulterated JPEG images from the untouched ones. Experimental results indicate that our method is very successful in detecting the presence of hidden data in JPEG steganography produced by CryptoBola, F5 algorithm, and JPHS. And the proposed feature set outperforms two other well-known feature sets: Histogram Characteristic Function Center Of Mass (HCFCOM), and the high-order moment statistical model in the multi-scale decomposition using wavelet-like transform.