DWT-SVD Based Image Watermarking

Abstract

Protecting digital multimedia content has become a major concern for content creators and service providers due to the rapid growth of digital media distribution. Digital watermarking has emerged as an effective technique for ensuring copyright protection and verifying ownership of multimedia data such as images, audio, and video. Numerous watermarking methods have been proposed to embed hidden information within multimedia files without affecting their visual or auditory quality.

Among the various techniques available, the Discrete Wavelet Transform (DWT) has gained significant attention because of its ability to represent image information in different frequency bands. In DWT-based watermarking approaches, watermark information is embedded by modifying the transform coefficients of the host image.

In this work, a hybrid watermarking technique that combines Discrete Wavelet Transform (DWT) and Singular Value Decomposition (SVD) is proposed. The cover image is first decomposed into four sub-bands using the DWT. After this decomposition, SVD is applied to each sub-band, and the watermark is embedded by modifying the singular values. By distributing the watermark information across multiple frequency components, the proposed method improves resistance against various signal processing attacks and enhances the overall robustness of the watermarking system.