In this paper, a robust digital audio content protection scheme is presented. The hidden data is chaotically modulated and embedded into integer modified discrete cosine transform (IntMDCT) frequency components to ensure the imperceptibility and security of the embedded watermark. The detection and extraction of hidden signals are explored using Independent Component Analysis (ICA) to exploit the statistical independence of components. These combined techniques provide a robust watermarking scheme with low distortion, large payload, high audio quality and full recovery. Experimental simulations have been performed on audio signals and show good robustness against common attacks with unnoticeable distortion.
|Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
|10th Pacific Rim Conference on Multimedia, PCM 2009
|15/12/09 → 18/12/09
- Audio Content Protection
- Theoretical Computer Science
- General Computer Science