Data Hiding Detection Based on DWT and Zernike Moments

In this paper, we present a data hiding detection (steganalysis) method based on Zernike moments of discrete wavelet coefficients of an image and SVM. We use Zernike moments of DWT subbands coefficients of an images as features for steganalysis and use support vector machines (SVM) for classification of stego and non-stego images. Experimental results show that there are different between these features for stago and non-stego images and these features are convenient for steganalysis. With randomly selected 400 images for training and the remaining 800 images for testing, the proposed steganalysis system can achieve a correct classification rate of 98.1% for Cox et al. algorithm, 99% for Corvi watermarking algorithm and 99.2% for Zhu algorithm. For combination of algorithms we reach to 95.8% correct detection rate.