A new fragile image-in-audio watermarking scheme for tamper detection

Watermarking, steganography and cryptography are used extensively to protect data that is being transmitted through the internet. Watermarking in particular is used to protect the copyright of multimedia, for tamper detection, for authentication, tracing illegal copies and as information carriers. They are used in images, audio, text, video and all kinds of multimedia. LSB (Least significant bit) water-marking technique has been used for fragile watermarking in the spatial domain. It is both easy to implement as well as effective. In this paper, a DCT-LSB (Discrete cosine transform) and a DST-LSB (Discrete sine transform) audio water-marking methods are proposed which show much higher level of sensitivities than LSB. PSNR (Peak signal to noise ratio) and NC (Normalized correlation) have been used to detect the imperceptibility of the embedded audio and tamper detection. The proposed methods are limited to tamper detection and do not per-form tamper localization. This has been implemented using Matlab.