Image Steganalysis Based on Statistical Moments of Wavelet Subband Histogram of Images with Least Significant Bit planes

This paper proposed a new image Steganalysis scheme based on statistical moments of histogram of multi-level wavelet subbands in frequency domain. These wavelet subbands derived from an image that has some of least significant bits of the grey level test image and some of its most significant bit planes are removed. Then we decompose the image using three-level Haar discrete wavelet transform (DWT) into 13 subbands (here the image itself is considered as the LL0 subband). The Fourier transform of each subband histogram, is calculated. The first three statistical moments of each subband histogram are selected to form a 39-dimensional feature vector for Steganalysis. Support Vector Machines (SVM) classifier is then used to discriminate between stego-images and innocent images. We experiment our proposed scheme on LSB, Cox and QIM data hiding method. Experimental results show that the proposed method improves the detection rate especially for LSB steganography.