Robust image watermarking scheme with transform domain technique on resisting different attacks: impact of self-improved optimization algorithm

With the rapid development of multimedia technology and its creation, the globe has seen an enormous rise in digital media attacks that replicate, tamper, and forge with data, resulting in infractions of intellectual property laws. Watermarking methods is a particular technique that addresses this issue and seeks to embed the knowledge in additional secured data in a distinctive way. However, implementing simply a transformation technique is insufficient to safeguard the image. In this paper, a novel robust image watermarking scheme is proposed with two processes: the embedding process and the extraction process. In the embedding process, the DWT level-2 decomposition is employed to retrieve fragmented sub-bands and decomposed images for both the cover image and the watermark image. Then these sub-bands are selected and further perform DWT for both images. Subsequently, the scrambling process is deployed in this work, in which the bitwise XOR operation is performed for the modified Arnold map and Gingerbread Man chaotic map that scrambles the image and retrieves the transformed coordinates image for the watermark image. Further, the encryption process takes place that encrypts the image with a key. Also, proposes the SIBWO (Self Improved Beluga Whale Optimization) algorithm to optimize the key, and then the encrypted image is embedded with the DWT cover image to produce a watermarked image. On the other hand, decryption on the watermarked image takes place then the decrypted image and the output of the Gingerbread Man chaotic map undergoes inverse bitwise XOR operation. Afterwards, the intermediate outcome from this operation is inputted into the inverse Modified Arnold map and then inverse DWT is applied to obtain the original watermark image. The proposed technique is evaluated against several types of attacks: Gaussian noise attack, Image rotation, and Image filtering. Experimental results indicate that the proposed watermarking technique is robust against these attacks while maintaining acceptable imperceptibility. An Ablation study on the proposed method is also performed which demonstrates that the proposed method is more efficient and secure.