High-capacity steganography: a global-adaptive-region discrete cosine transform approach

An increasing number of spatial and frequency domain data hiding techniques have been proposed to address the relatively low embedding capacities of image-based steganography. These techniques have brought promise of higher embedding capacities, albeit at the expense of lower perceptibility. This work proposes a new discrete cosine transform (DCT) approach for color image steganography and implements a global-adaptive-region (GAR) embedding scheme that allows for extremely high embedding capacities while maintaining enhanced perceptibility. The idea is to adapt the variable region size, used to hide the data, in each DCT block of the cover image to the amount of correlation of the image values in the corresponding block. We will demonstrate how this new technique achieves enhanced hiding capacities and perceptibility compared to other spatial, Fourier, and adaptive-region DCT based steganography schemes.