Title :
Information embedding codes matched to locally stationary Gaussian image models
Author :
Mihçak, M. Kivanç ; Moulin, Pierre
Author_Institution :
Microsoft Res., Redmond, WA, USA
Abstract :
We have designed a practical information embedding code for images. The method assumes a locally i.i.d. (independent identically distributed) Gaussian source model for wavelet image coefficients. The proposed algorithm is developed based on this model and recent information theoretic results on information hiding for parallel Gaussian channels. Promising results have been obtained on test images.
Keywords :
Gaussian channels; Gaussian distribution; combined source-channel coding; data encapsulation; image coding; transform coding; wavelet transforms; code matching; independent identically distributed source model; information embedding codes; information hiding; locally iid Gaussian source model; locally stationary Gaussian image models; parallel Gaussian channels; wavelet image coefficients; Data encapsulation; Decoding; Fingerprint recognition; Gaussian channels; Image coding; Robustness; Steganography; Testing; Tin; Watermarking;
Conference_Titel :
Image Processing. 2002. Proceedings. 2002 International Conference on
Print_ISBN :
0-7803-7622-6
DOI :
10.1109/ICIP.2002.1039906
