Block Based Image Steganography in Spatial and Frequency Domain

Abstract

Steganography is the art of hiding a secret message in different kind of multimedia (image, voice or video), such that the secret message is not detectable. In this paper, we propose two new algorithms, first one uses the spatial domain for steganography, where the host image is converted into blocks of bit-planes to insert the secret information. The algorithm divides the image into 8 bit planes and then the bit planes are further divided in to N x N blocks. The hidden message is inserted based on a chaotic sequence. We intend to find the most optimum bit plane to insert the hidden information, keeping high imperceptibility in terms of the human visual systems. The algorithm shows relatively good Mean Structural Similarity and Peak Signal to Noise Ratio values. The second algorithm is applied in the frequency domain where the host image is converted using the discrete wavelet transform. Then at second and third level of the transform, the secret information is inserted. The proposed algorithm divides wavelet level divide in to M x M blocks. The hidden message is inserted based on chaotic sequence in to the blocks. This algorithm shows better imperceptivity in terms of the human visual system and PSNR.