A watermarking scheme for color images that achieves optimality using the Transit Search Algorithm
Mahdi Tamimi, Ahmed Bencherqui, Mohamed Amine Tahiri, Hicham Karmouni, A. El Mloufy, Hassan Qjidaa, Mhamed Sayyouri · Mathematical Modeling and Computing · 2024
This paper describes an innovative watermarking method that combines discrete wavelet transform (DWT), Hessenberg decomposition (HD), and singular value decomposition (SVD). To do this, the main image and the watermark are divided into three channels (red, green and blue – RGB). Then, each part of the main image individually undergoes the steps of DWT, HD and SVD, while the watermark components are processed by SVD. Insertion of the watermark is carried out by adjusting the singular values of the watermark and the main image, using a watermark scaling factor (α). The optimal choice of α poses a challenge, so the transit search algorithm is employed to find a trade-off between visibility and robustness. To evaluate this method, comparisons are made with other studies using various optimization algorithms such as particle swarm optimization, artificial bee colony and fly optimization algorithm. The results of the experiments confirm the effectiveness of this technique.