Research On Reversible Data Hiding Techniques

The problem of reversible data hiding technology was posted in late1990s, and from then on, many novel reversible data hiding schemes had been proposed. As it is used in a wide range of applications about people’s production and living, such as intellectual property (IPR) protection, the remote medical diagnosis, judicial expertise, military covert communication, and so on. Image reversible data hiding is still a popular research topic in the area of image processing at the present time. The reversible data hiding is an important part of image watermarking.In this thesis, we not only give an overview of the recently proposed reversible data hiding techniques, but also devise a new scheme called reversible data hiding using locally linear Prediction (LLP), to deal with problems of embedding distortion and embedding capacity. Because there is a strong correlation between the neighbouring pixels, so we predict the pixel value according to its surrounding pixels. At the same time, we are inspired by locally linear embedding (LLE), which is an unsupervised learning algorithm that computes low-dimensional, neighborhood preserving embeddings of high-dimensional inputs. It first calculate the weighted coefficients of unknown pixels’s four nearest neighbours by least square method. Then it derive the prediction value through a linear combination of these four pixels. Experimental results show that the proposed method outperforms typical state-of-the-art algorithms in terms of embedding capacity and distortion by a large margin.