발간년도 : [2024]
| 논문정보 |
|
| 논문명(한글) |
[Vol.19, No.1] Wavelet Packet Based Watermarking Using Hessenberg and Singular Value Decomposition |
|
| 논문투고자 |
Su-Young Han |
|
| 논문내용 |
In this paper, we propose a wavelet packet based image watermarking algorithm using Hessenberg decomposition and singular value decomposition. The wavelet packet transform can properly embed the watermark in the sub-band excluding the lowest and highest frequencies, which can simultaneously improve the invisibility and robustness of the watermarking algorithm. In addition, Hessenberg decomposition and singular value decomposition is used to change the wavelet packet coefficients and watermark to increase the robustness. These matrix decomposition can enhance the robustness against geometric attacks, which other frequency domain based watermarking algorithms are vulnerable to. The invisibility and robustness of the algorithm were analyzed through simulations, and the PSNR of the watermarked image and the correlation between the extracted watermark and the original watermark were analyzed. All simulations were performed using MATLAB version R203b software. As a result, the PSNR of the embedded watermarked image is over 38 dB, which is sufficiently invisible and shows high performance compared to wavelet based watermarking. The correlation between the embedded watermark and the original watermark is more than 0.99, which shows that it is well maintained under various attacks. In particular, the simulation shows that the improvement is greater for aerial images, which are more complex and contain relatively more high-frequency regions. |
|
| 첨부논문 |
|
|
|
|
|
자료실 > 논문지 
