An Effective Hybrid Image Watermarking Encrypted and Decrypted Technique for Copyright Protection

Image watermarking serves as a visual deterrent, signaling potential infringers that the image is protected by copyright. While various watermarking methods have been developed in the past, existing models encounter limitations such as a lack of suitability for diverse data types and ineffective data generalization. Additionally, the absence of standardization in watermarking techniques poses a significant challenge. In this proposed approach, images are input into an improved discrete Fourier transform (IDFT) to convert signals from time domain. The Arnold Transform is initially applied to convert images into a signal format. Subsequently, Walsh-Hadamard Transform (WHT) is employed for non-sinusoidal, orthogonal signal transformation. Finally, Singular Value Decomposition (SVD) is implemented to decrease number of dimensions in high-dimensional data, facilitating easier visualization and analysis. The average Peak Signal-to-Noise Ratio (PSNR), Normalized Cross-Correlation (NCC), and Bit Error Rate (BER) of the hybrid IDFT-WHT-SVD watermarking scheme are reported as 56.58, 0.2651, and 0.9902, respectively. This proposed model underwent comparison with other existing models, including CNN-based robust watermarking, Hyper-Chaotically Encrypted, and Cryptographic Technique.