A Robust Image Watermarking Scheme Based on Discrete Shearlet Transform (DST) and Singular Value Decomposition (SVD)

A robust image watermarking scheme based on Discrete Shearlet Transform (DST) and Singular Value Decomposition is presented in this paper. The multiscale directional representation feature of DST is used to provide a high degree of imperceptibility, while Singular Value decomposition improves robustness and security. The proposed scheme utilises the quantisation index modulation and Arnold transformation. The scheme employs a block-based technique for watermark embedding which enhances its imperceptibility and security. Several experiments were conducted using standard test images to ascertain the performance of the scheme against common image processing and geo-metric attacks. The correlation coefficients of the original and watermarked test images show that the proposed scheme has excellent imperceptibility and robust-ness characteristics. Hence, this approach can be used for various real-time applications, such as copyright protection and broadcast monitoring.