Reversible Data Hiding with Prediction-Based Multi-Bit Embedding and Multi-Level Difference Alteration

Applications of digital rights management (DRM) have been developed for some years. Among them, watermarking or data hiding techniques work along with encryption algorithms to protect the copyright of multimedia contents. Nowadays, people tend to capture images with smart phones, and then they can transmit the image online instantly. Therefore, DRM relating issues have practical applications, and methods with reversible data hiding have also been developed. Among the categories in reversible data hiding, by altering the difference values relating to original image seems effective with good performance. We employ prediction techniques to produce the predicted image, and then to calculate the differences between original image and predicted one. Unlike conventional methods to embed one bit at a time, we embed multiple bits at a time based on the characteristics of images to look for enhanced amount of capacity. Simulation results have pointed out the enhancement in embedding capacities, and the improvements in overall performances over relating methods in literature.