The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Robustesse et invisibilité sont deux contraintes contraires pour un tatouage invisible robuste. Au lieu de la stratégie conventionnelle avec le modèle du système visuel humain (HVS), cet article présente une approche adaptative au contenu pour optimiser davantage la contrainte entre eux. Pour atteindre cet objectif, les mesures basées sur l'entropie et basées sur le HVS intégré (IHVS) sont construites de manière à choisir de manière adaptative les composants appropriés pour l'insertion et la détection de filigrane. Un tel type de schéma donne potentiellement lieu à un problème de synchronisation entre le codeur et le décodeur dans le cadre d'un tatouage aveugle, qui est ensuite résolu en incorporant le code de répétition-accumulation (RA) avec effacement et correction d'erreurs. De plus, un nouveau détecteur basé sur un modèle de Markov caché (HMM) dans le domaine des ondelettes est introduit pour réduire la complexité de calcul et est développé ultérieurement pour éviter la transmission des paramètres HMM avec seulement un léger sacrifice des performances de détection. Les résultats expérimentaux montrent que l’algorithme proposé peut obtenir une amélioration considérable des performances de robustesse avec la même distorsion que l’algorithme traditionnel.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copier
Chuntao WANG, Jiangqun NI, Rongyue ZHANG, Goo-Rak KWON, Sung-Jea KO, "Content-Adaptive Robust Image Watermarking with Posterior HMM-Based Detector" in IEICE TRANSACTIONS on Fundamentals,
vol. E91-A, no. 8, pp. 1953-1960, August 2008, doi: 10.1093/ietfec/e91-a.8.1953.
Abstract: Robustness and invisibility are two contrary constraints for robust invisible watermarking. Instead of the conventional strategy with human visual system (HVS) model, this paper presents a content-adaptive approach to further optimize the constraint between them. To reach this target, the entropy-based and integrated HVS (IHVS) based measures are constructed so as to adaptively choose the suitable components for watermark insertion and detection. Such a kind of scheme potentially gives rise to synchronization problem between the encoder and decoder under the framework of blind watermarking, which is then solved by incorporating the repeat-accumulate (RA) code with erasure and error correction. Moreover, a new hidden Markov model (HMM) based detector in wavelet domain is introduced to reduce the computation complexity and is further developed into a posterior one to avoid the transmission of HMM parameters with only a little sacrifice of detection performance. Experimental results show that the proposed algorithm can obtain considerable improvement in robustness performance with the same distortion as the traditional one.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1093/ietfec/e91-a.8.1953/_p
Copier
@ARTICLE{e91-a_8_1953,
author={Chuntao WANG, Jiangqun NI, Rongyue ZHANG, Goo-Rak KWON, Sung-Jea KO, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Content-Adaptive Robust Image Watermarking with Posterior HMM-Based Detector},
year={2008},
volume={E91-A},
number={8},
pages={1953-1960},
abstract={Robustness and invisibility are two contrary constraints for robust invisible watermarking. Instead of the conventional strategy with human visual system (HVS) model, this paper presents a content-adaptive approach to further optimize the constraint between them. To reach this target, the entropy-based and integrated HVS (IHVS) based measures are constructed so as to adaptively choose the suitable components for watermark insertion and detection. Such a kind of scheme potentially gives rise to synchronization problem between the encoder and decoder under the framework of blind watermarking, which is then solved by incorporating the repeat-accumulate (RA) code with erasure and error correction. Moreover, a new hidden Markov model (HMM) based detector in wavelet domain is introduced to reduce the computation complexity and is further developed into a posterior one to avoid the transmission of HMM parameters with only a little sacrifice of detection performance. Experimental results show that the proposed algorithm can obtain considerable improvement in robustness performance with the same distortion as the traditional one.},
keywords={},
doi={10.1093/ietfec/e91-a.8.1953},
ISSN={1745-1337},
month={August},}
Copier
TY - JOUR
TI - Content-Adaptive Robust Image Watermarking with Posterior HMM-Based Detector
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1953
EP - 1960
AU - Chuntao WANG
AU - Jiangqun NI
AU - Rongyue ZHANG
AU - Goo-Rak KWON
AU - Sung-Jea KO
PY - 2008
DO - 10.1093/ietfec/e91-a.8.1953
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E91-A
IS - 8
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - August 2008
AB - Robustness and invisibility are two contrary constraints for robust invisible watermarking. Instead of the conventional strategy with human visual system (HVS) model, this paper presents a content-adaptive approach to further optimize the constraint between them. To reach this target, the entropy-based and integrated HVS (IHVS) based measures are constructed so as to adaptively choose the suitable components for watermark insertion and detection. Such a kind of scheme potentially gives rise to synchronization problem between the encoder and decoder under the framework of blind watermarking, which is then solved by incorporating the repeat-accumulate (RA) code with erasure and error correction. Moreover, a new hidden Markov model (HMM) based detector in wavelet domain is introduced to reduce the computation complexity and is further developed into a posterior one to avoid the transmission of HMM parameters with only a little sacrifice of detection performance. Experimental results show that the proposed algorithm can obtain considerable improvement in robustness performance with the same distortion as the traditional one.
ER -