Semi-reversible quantization based data hiding using missing samples recovery technique

A blind semi-invertible quantization based data hiding scheme, which reconstructs the original signal with high precision has been proposed. In order to produce correlated quantization yielding reversibility of the quantization based approach, a new transform domain has been introduced. In decoder, by compensating the quantization error and using the iterative technique, the original signal is recovered; then the watermarked signal is compared with the reconstructed original signal and hidden data is retrieved. Simulation results show that the proposed method in comparison with other reversible methods imposes less distortion and thus a higher Signal to Noise Ratio (SNR) is achieved.