Digital Copyright Protection of Images Based on Invasive Weed Optimisation Algorithm

Research on digital copyright protection for images provides digital media platforms and social media with important tools to ensure that users do not infringe on the intellectual property rights of others when sharing and distributing digital images. This contributes to a fairer and more legitimate digital ecosystem that encourages creators to actively share their work while protecting their creative output. Therefore, the digital copyright protection technique for images with invasive weed algorithm is proposed. Firstly, the principles and characteristics of human visual system are introduced, and the perceptual characteristics of colour images and the conversion relationship between colour spaces are analysed. Then, the proposed watermarking algorithm is described in detail, in which the watermarked image is disrupted to increase the security of the algorithm. An invasive weed optimisation algorithm is used to select the appropriate embedding strength to embed the watermark information into the carrier image wavelet transform coefficients to balance the algorithm robustness and invisibility. The optimal solution of embedding strength is obtained by seed dispersal and weed multiple competitive optimisation. Finally, experiments were conducted using MATLAB software and the results were analysed. Experimental results show that compared with the common particle swarm optimization and differential evolution algorithm, the proposed method not only improves the applicability of the algorithm, but also shows strong robustness to noise attacks, compression and geometric attacks such as rotation, shearing and translation on the basis of ensuring the invisibility of the algorithm. © 2024, Taiwan Ubiquitous Information CO LTD. All rights reserved.