Discrete Wavelet Transform Based Watermarking Using Modified Matrix Encoding

Conventional Robust Reversible Watermarking methods have limited robustness in extracting watermarks from the watermarked images destroyed by different unintentional attacks and some of them suffer from extremely poor invisibility for water marked images. It is necessary to have a framework to address these three problems and further improve its performance. It presents a novel practical structure, wavelet-area arithmetical quantity histogram shifting and cluster (WSQH-SC). Compared with conservative methods, WSQH-SC resourcefully constructs new watermark  in addition to removal procedures by histogram shifting and clustering, which are important for civilizing robustness and reducing run-time difficulty in addition, WSQH-SC includes the property-inspired pixel adjustment to effectively handle overflow and underflow of pixels. This results in acceptable reversibility and invisibility. To increase its practical applicability WSQH-SC designs an enhanced pixel-wise masking to balance robustness and invisibility. It perform extensive experiments over normal, medicinal and artificial opening radar imagery to show the effectiveness of WSQH-SC by comparing with the histogram rotation-based and histogram distribution constrained methods.