Steganography and watermarking of digital images using singular value decomposition
Saibal Kumar Pal, P.K. Saxena, Sunil Kumar Muttoo · Journal of Discrete Mathematical Sciences and Cryptography · 2005
Singular Value decomposition (SVD), a scheme for matrix diagonalization, has shown promising results in the area of digital image compression. An image matrix is approximated with a rank lower than that of the original image without affecting its perceptual quality. This paper addresses two important aspects of information hiding using SVD namely steganography and watermarking. The lesser significant singular values are used to encode secret messages and carry out hidden communication. Watermarking schemes are designed by slightly modifying the most significant singular values for encoding vital data in the media. By manipulating singular values in different ways, we show the possibility of high-capacity steganography as well as robust watermarking schemes with the ability to resist attacks by active wardens monitoring the communication channels.