Increasing the Performance of Difference Expansion-based Steganography when Securing Medical Data

Abstract

Steganograpy has been widely used for hiding secret data in various types of other data, such as images, audio, or video. In order to obtain this secret data, an extraction algorithm is used according to the respective steganographic embedding algorithm. Nevertheless, many data hiding methods have three common problems: reversibility or obtaining both secret and cover data in the extraction process, similarity between cover data and stego data, and the size of the secret data that can be hidden. In this paper, we focus on the last two problems while still maintaining reversibility (the first problem) by proposing four methods which are developed based on Difference Expansion. In the first three methods, we use a threshold to control the stego image quality, while in the fourth we use the smoothness level instead of a threshold value. In these proposed methods, we use text and images as the secret and cover data, respectively, where the secret data is embedded into the difference of pixel blocks. The size of these blocks varies. In addition, the difference is reduced before the secret data is embedded. The experimental results show that the secret data as well as the cover data can be reconstructed to be exactly the same as the original. Furthermore, both the capacity of the secret data and the similarity level increase for certain block sizes.