Steganalysis based on Markov Model of Thresholded Prediction-Error Image

摘要: A steganalysis system based on 2-D Markov chain of thresholded prediction-error image is proposed in this paper. Image pixels are predicted with their neighboring pixels, and the generated by subtracting prediction value from pixel then a predefined threshold. The empirical transition matrixes along horizontal, vertical diagonal directions serve as features for steganalysis. Support vector machines (SVM) utilized classifier. effectiveness has been demonstrated extensive experimental investigation. detection rate Cox et al.' s non-blind spread spectrum (SS) data hiding method, Piva blind SS generic QIM method (as embedding being 0.1 bpp (bits per pixel)) all above 90% over an database consisting approximately 4000 images. For LSB (with various rates), our achieves 85% above.