Feature-Based Robust Image Watermarking Using DTT and SVD for Copyright Protection.

Shivani, J. L. Divya; Senapati, Ranjan K.

This paper presents a simple yet efficient watermarking algorithm based on scale invariant feature transform (SIFT), singular value decomposition (SVD) and discrete Tchebichef transform (DTT). The generation process of SIFT algorithm, and generation of feature points are described. In the proposed method, first DTT is applied to the 8×8 blocks of an image data. The 8×8 blocks of coefficients are then arranged like coefficients in a 3 level wavelet pyramid. SVD is applied to appropriate pyramid level/s for embedding the watermark. The feature points obtained from SIFT are used for correction of geometrical attacks, e.g., rotation, scale and translation (RST) attacks. The effect of combined attacks such as rotation and scale, translation and scale, translation and rotation, rotation with noise, translation with noise, scale with noise, rotation with JPEG compression have been also analysed. Experimental results demonstrate that the proposed algorithm has stronger robustness and better imperceptibility compared to previous schemes in most of the attacks.

SINGULAR value decomposition; ALGORITHMS; DIGITAL watermarking; PREVENTION of copyright infringement

International Journal of Simulation: Systems, Science & Technology, 2018, Vol 19, Issue 6, p1

1473-8031

Academic Journal

10.5013/IJSSST.a.19.06.07