Blind prediction-based wavelet watermarking

Abstract

In recent years, wavelet transform has been widely applied in multimedia signal processing and digital watermarking is involved for ensuring security. This study presents a blind wavelet-based watermarking incorporated with the Human Visual System (HVS), which embeds watermarks into detail-subband coefficients. Since imperceptibility is the most significant issue in watermarking, the approximation band is maintained constant, while the detail subbands are modified to carry information. The perceptual embedded weights for all subbands are determined according to the Just Noticeable Distortion (JND) criterion. The strength of the modification is investigated to provide a balanced result between robustness and image quality. In the decoder, the Least-Mean-Square (LMS) is employed to predict the original detail-subband coefficients and then extract the embedded watermarks. As documented in the experimental results, the proposed method provides good robustness and excellent image quality.