Research Article Open Access

Extracting Spread-Spectrum Hidden Data from Digital Media

Abstract

We consider the problem of extracting blindly data embedded over a wide band in a spectrum (transform) domain of a digital medium (image, audio, video). We develop a novel multi-carrier/signature iterative generalized least-squares (M-IGLS) core procedure to seek unknown data hidden in hosts via multi-carrier spreadspectrum embedding. Neither the original host nor the embedding carriers are assumed available. Experimental studies on images show that the developed algorithm can achieve recovery probability of error close to what may be attained with known embedding carriers and host autocorrelation matrix.

Jyothi B. Rameshkumar H K

To read the full article Download Full Article