.png){kind=icon}
.png){kind=icon}
Return
UDTCWT difference domain statistical decoder using vector-based Weibull PDF
.png){kind=icon}
PRE
AI
DOI：10.1007/s11042-022-13229-9.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Invisibility, data payload and robustness are widely regarded as the three main attributes that are critical to any digital image watermarking schemes. They restrict each other, so it is a research focus to strike a favorable balance between them. Multiplicative watermarking based on statistical models is an effective method to achieve the trade-off between these three demands. However, most of the existing approaches until now focus on employing transform coefficients, which are usually poor in resistance to multiple attacks. We proposed a new digital image watermarking system in the difference domain of Undecimated Dual Tree Complex Wavelet Transform (UDTCWT) in this article, in which the UDTCWT difference coefficients are used for hiding the watermark information and constructing the watermark decoder. In the process of embedding the watermark, the watermark message is hidden in the UDTCWT difference domain of the carrier image using a multiplicative manner so as to achieve better imperceptibility and robustness. In watermark decoding, we first prove that the vector based Weibull distribution can suitably fit the distribution of the UDTCWT difference coefficients. And then, we constructed a blind statistical digital image watermark decoder in UDTCWT difference coefficients, combining the vector based Weibull distribution with the locally most powerful (LMP) rule. Experimental results on some standard test images demonstrate the efficacy and superiority of the proposed approach.
Keywords:
Statistical image watermarking
UDTCWT difference coefficients
Vector based Weibull distribution
MAP parameters estimation
LMP test criterion
Journal
IF:
3
Papers:
1.9W
Citations:
3.2W