Video watermarking with Curvelet transform

Abstract

In wireless communications, secured transmission of video has gained considerable research interest. The traditional wavelets are poor in handling curve singularities. Therefore, to handle them curvelet transforms are used. This transform exhibits minimum mean square error between the original and reconstructed image. Therefore, in this work, a watermarking is proposed with this transform using quantization index modulation. The watermark is embedded in the low resolution part of curvelet coefficients using dither quantization. To provide security, the mark is enciphered with Toral Amorphism. The algorithm is blind and the experimental results confirms that the proposed watermarking scheme offers high imperceptibility at the same time robust to attacks such as rotation, blurring due to camera motion and H.264 compression. To evaluate the performance of the watermarking scheme performance metrics Peak Signal to Noise ratio (PSNR) and Normalized Cross Correlation (NCC) are used in the work. The proposed algorithm is compared with recent works and the proposed algorithm offers high invisibility and resistance to attacks.