LWT-DCT-SVD and DWT-DCT-SVD based watermarking schemes with their performance enhancement using Jaya and Particle swarm optimization and comparison of results under various attacks

Robust and undetectable watermarking scheme is important for copyright protection. This paper proposes a DWT-DCT-SVD and LWT-DCT-SVD based image watermarking technology, as well as a comparison of the DWT and LWT systems. The host picture is first divided into sub-bands throughout the embedding phase using multilevel DWT or LWT. The image is divided into various high and low-frequency sub-bands using DWT. LWT breaks up the original filters into a series of smaller structures and provides a faster algorithm. The ensuing sub-bands area unit is then used as the input for DCT. The SVD is performed after applying DCT. The watermark is then inserted into the input image sub-band by using particle swarm optimization intelligence and JAYA optimization to discover an acceptable scale factor. The proposed methodology is compared with other methods under a variety of first-generation and second-generation attacks, including filtering, noise, JPEG compression, JPEG2000 compression, sharpening, thresholding, dithering, motion blur, and shear attack. Furthermore, the proposed watermarking system has a reasonable level of robustness and imperceptibility in the face of most attacks, therefore it is a suitable technique for Copywrite protection. This paper proposes two watermarking schemes based on discrete wavelet transform (DWT) namely DWT-DCT-SVD and lifting wavelet transform (LWT) based scheme namely LWT-DCT-SVD. The comparison of these two schemes is also carried out.