Robust and blind watermarking algorithm based on DCT and SVD in the contourlet domain
Cheng Tian,Ru-Hong Wen,Wei-Ping Zou,Li-Hua Gong
DOI: https://doi.org/10.1007/s11042-019-08530-z
IF: 2.577
2019-12-23
Multimedia Tools and Applications
Abstract:A blind watermarking algorithm in multiple transform domains is presented for copyright protection. This robust algorithm is designed by fusing contourlet transform (CT), discrete cosine transform (DCT) and singular value decomposition (SVD). The host image is first decomposed by one-level CT and its low frequency sub-band is partitioned into 8 × 8 non-overlapping blocks. Then, each block is transformed by DCT and several middle frequency DCT coefficients with good stability are selected to construct the carrier matrix. Finally, the watermark is embedded by modifying the largest singular values of two carrier matrices. Besides, the geometric distortion factor is estimated with the speed up robust features (SURF) algorithm. The proposed watermarking scheme is evaluated in terms of imperceptibility and robustness. Experimental results demonstrate that the proposed watermarking scheme preforms better in terms of invisibility and robustness than other related schemes.
computer science, information systems, theory & methods,engineering, electrical & electronic, software engineering