A Reversible Watermarking Algorithm Based on Four-Neighbors Context Prediction for Tongue Images

Abstract

Reversible watermarking has been used in many sensitive digital images, such as medical and military images, because of the ability to embed data with zero loss of host information. This paper describes a reversible watermark embedding algorithm, based on context prediction error for color tongue images, that can achieve high data embedding rates and effective copyright protection for such images. The algorithm exploits the high correlation that is inherent among the neighboring pixels of tongue images, forming a more effective method for embedding data by using the proposed prediction and extended difference-expansion techniques in different components. When extracting, the algorithm can restore the original tongue image accurately, meeting the requirements required for medical images. The experimental results show that the algorithm is easy to implement and that it achieves very high data embedding rates while keeping the resulting distortion low.