A NOVEL IMAGE WATERMARKING SYSTEM UTILIZING HMM IN WAVELET DOMAIN

摘要

Achieving high capacity, sufficient imperceptibility and enough robustness is crucial for a good image watermarking scheme. In this paper a novel image watermarking algorithm is proposed which uses Taylor series approximated locally optimum test (TSLOT) detector, Hidden Markov Model (HMM) in wavelet domain using bi-orthogonal 9/7 wavelet TSLOT based HMM in wavelet domain is established to handle the question of unavailability of exact embedding strength in the receiver due to informed embedding. Then based on the TSLOT detector new HMM-based spherical codes are built to afford an effective balance among robustness and distortion. The process of informed embedding is expressed as an optimization problem under the robustness and distortion restrictions and the genetic algorithm (GA) is then employed to solve this problem. Simulation results demonstrate that the proposed informed watermarking algorithm has high robustness against common attacks in signal processing.