Robust Digital Watermarking for Colour Images Using GBT & SVD

Abstract: Copyright infringement poses a significant challenge in the creative industry, particularly with the exponential growth of digital content. Digital watermarking is a promising technique to mitigate this problem. Its primary objective is to embed an imperceptible digital watermark into digital content, such as images or videos, to protect against unauthorized use or reproduction. This paper presents a novel hybrid approach for digital watermarking that addresses the limitations of existing methods, namely the Discrete-Haar Wavelet Transform (DHWT) and Least Significant Bit (LSB) steganography. The proposed approach integrates Graph Based Transform (GBT) and Singular Value Decomposition (SVD) for data watermarking, while incorporating a scrambling operation to enhance the security of the watermark. Compared to DHWT, the proposed approach has lower computational cost and reduced distortion near image or video edges. In contrast to LSB, it utilizes transform domain techniques resulting in a more robust aftermath. The experimental results demonstrate the effectiveness of the proposed approach, indicating its potential as a viable solution for digital watermarking.

Keywords: DHWT, digital watermarking, LSB, SVD, GBT, scrambling