Blind Steganography in Color Images by Double Wavelet Transform and Improved Arnold Transform

Abstract

Steganography is a method which can put data into a media without a tangible impact on the cover media. In addition, the hidden data can be extracted with minimal differences. In this paper, two-dimensional discrete wavelet transform is used for steganography in 24-bit color images. This steganography is of blind type that has no need for original images to extract the secret image. In this algorithm, by the help of a structural similarity (SSIM) and a two-dimensional correlation coefficient, it is tried to select part of sub-band cover image instead of embedding location. These sub-bands are obtained by 3-levels of applying the DWT. Also to increase the steganography resistance against cropping or insert visible watermark, two channels of color image is used simultaneously. In order to raise the security, an encryption algorithm based on Arnold transform was also added to the steganography operation.