New watermarking algorithm utilizing quaternion Fourier transform with advanced scrambling and secure encryption

• Published in: Multimedia tools and applications 2021-04, Vol.80 (9), p.13367-13387
• Main Authors: Bhatti, Uzair Aslam, Yuan, Linwang, Yu, Zhaoyuan, Li, JingBing, Nawaz, Saqib Ali, Mehmood, Anum, Zhang, Kun
• Format: Article
• Language: English
• Subjects: Algorithms; Color imagery; Computer Communication Networks; Computer Science; Data Structures and Information Theory; Digital watermarking; Embedding; Encryption; Fourier transforms; Gray scale; Image processing; Mathematical analysis; Multimedia; Multimedia Information Systems; Quaternions; Scrambling (communication); Security; Signal to noise ratio; Special Purpose and Application-Based Systems; Vectors (mathematics)
• ISSN: 1380-7501; 1573-7721
• DOI: 10.1007/s11042-020-10257-1

Digital watermarking technology, as a powerful tool for copyright protection and content authentication of multimedia works, has received increasing attention, and the current image watermarking technology has developed significantly. Generally, embedding a watermark is done in grayscale images, mainly due to the fact that grayscale images are easier to process than color images, and grayscale images only contain brightness information and color-free information, in which an embedded watermark will not produce new color components. In real life, however, color images are more common than grayscale image information. To improve the security of the watermark information and the ability to embed the location and improve the security of the algorithm against a variety of attacks, this paper proposes an algorithm based on quaternion Fourier transform (QFT) with chaotic encryption and Arnold scrambling. Geometric Algebra (GA) can process color images in the form of vectors, and each component of Red,Green and Blue (RGB) should be processed separately. We utilized quaternion (which is a subalgebra of GA) and effectively completed color image processing by utilizing Fourier transform. After utilizing QFT, each component was made more secure by scrambling the pixels of the watermark and performing encryption utilizing chaotic sequencing. Different types of results utilizing Mean Square Error (MSE),Peak Signal to Noise Ratio (PSNR),Structure Similiarity Index Metric (SSIM), etc. were computed to verify the performance of our proposed algorithm and show that the robustness and security are high by utilizing our method.