A Comparative Study of Image-In-Image Steganography Using Three Methods of Least Significant Bit, Discrete Wavelet Transform and Singular Value Decomposition

Abstract

Today, with the growth in application of communication equipment as well as the use of Internet for information exchange, the need for data security and protection is felt more than before. Various methods have been proposed in recent years to realize this goal. One efficient, widely used method is hiding of information within data of insignificant value. An image, due to its large amount of information, may embed different data within, hence providing a decent safety margin for critical data. Three methods, including least significant bit (LSB), discrete wavelet transform (DWT) and singular value decomposition (SVD) were used in this study for image-in-image steganography. Each method has advantages and disadvantages. Advantages of LSB include high capacity and simplicity, however, as for the disadvantages, it is vulnerable against attacks such as cropping, salt and pepper noise, and compression. Therefore, the other two methods, DWT and SVD, which are robust against the mentioned attack, are used as steganography in frequency domain. Among the methods used in this study, SVD least affects the host image, making it more suitable as compared with its other two counterparts.