A new robust image adaptive steganography method in wavelet domain

A new robust adaptive data hiding method in the wavelet transform domain of an image is introduced in this paper. The proposed steganography method embeds the secret data in the blocks of an image that seems to be noisy based on the bit plane complexity of each block and does not destroy the co-occurrence matrix of wavelet coefficient. Most of the steganalysis methods are based on the moments of histogram or co-occurrence matrix of an image. Since the histogram is one-dimensional case of co-occurrence matrix, we embed the data in blocks which causes minimal changes to the co-occurrence matrix. In our method we use the one-third and rounding methods for embedding data in wavelet coefficients, and retain the co-occurrence matrix of wavelet coefficient. Our method has more robustness in comparison with other steganography methods in the same capacity.