Efficient Steganography Techniques for Secure Data Transfer using Graph Signal Processing

Abstract

Data communication through the Internet offers ease in digital communica- newlinetion. Transferring multimedia content over the public network is a challenge newlineto its security, privacy, right of ownership, authentication, and confidentiality. newlineSteganography came up as one of the methods to transfer secret multimedia newlinecontent securely over the public network. It is the process of concealing sensi- newlinetive information within the cover medium by hiding the existence or presence newlineof the data to be transmitted. In the steganography technique, imperceptibil- newlineity, embedding capacity, and robustness are major concerns as they deal with newlinethe visual quality of the stego image and resistance against image processing newlineattacks respectively. Designing a new steganography algorithm can introduce newlineimprovements in the embedding process, key management technique, encryp- newlinetion, and scrambling mechanisms. These features can help to protect the newlineconfidentiality, integrity, and authenticity of the embedded data by maintain- newlineing the visual quality. Although steganography approaches have improved newlineover the years, they still have a number of shortcomings, including detectabil- newlineity, capacity, robustness, attack vulnerability, and complexity. While keeping newlinethese concerns in mind, we designed and developed three efficient and robust newlinesteganography techniques to transfer multimedia data. newlineResearchers generally used some traditional and standard transform domain newlinetechniques to transform the image before embedding the data. In this the- newlinesis, an emerging area of graph signal processing (GSP) in the field of image newlinesteganography is explored to transform the image into a graph signal to en- newlinehance the performance of the techniques. newlineTo improve imperceptibility and robustness, an efficient and secure image newlinesteganography technique based on graph signal processing is presented in this newlinethesis. The scrambling of a secret image using Arnold cat map transform be- newlinefore embedding improves the security of the concealed image.