Robust watermarking for medical images based on Arnold scrambling and DWT-DFT

Currently, digital information management system has played an increasingly important role in modern health care system. However, the Information security issues are exposed gradually. When an exception occurs, digital medical image, as a diagnostic basis, can not serve as the evidence of medical accident case, and may result in delayed diagnosing or even misdiagnosing for patients. In order to solve this problem, this paper proposes an algorithm that employs the Arnold scrambling to preprocess on original watermarking, such that the watermarked medical image technology has security under the cryptography sense, for double protection. Moreover, the algorithm combines the image visual feature vector, encryption technology and the concept of third party, without the sophisticated process to finding the Region of Interest (ROI) of medical images. Therefore, watermarking can be extracted and embedded more faster. The experimental results show that the algorithm has satisfactory robustness and invisibility under conventional attacks and geometrical attacks in different levels.