Robust Encrypted Watermarking for Medical Images Based on DWT-DCT and Tent Mapping in Encrypted Domain

Abstract

Medical images are a special kind of image, since the information contained within must be kept strictly confidential. With a view to protecting personal information, a plethora of research has been carried out in the field of digital watermarking. However, in previous studies, watermarkings were encrypted and the processing of original medical images was unencrypted. Original medical images also contain a large amount of patient information that needs to be kept confidential; therefore, the present paper proposes a robust watermarking algorithm for medical images based on DWT-DCT and a tent map in an encrypted domain. Discrete Cosine Transform (DCT) is a form of Fourier transform, and only uses the actual numbers. Discrete Wavelet Transform (DWT) is a type of signal processing tool that is used to discretize the scale and displacement of basic wavele. In the present paper, DWT and DCT are combined to increase robustness. Firstly, the original medical image and the watermarking were encrypted by chaotic tent mapping. The feature vector of the encrypted medical image was extracted by DWT-DCT, and the watermarking was subsequently embedded and extracted. Finally, the normalized correlation coefficient (NC value) between the embedded and extracted watermarking was calculated to observe its robustness. The experimental results show that this algorithm is robust against conventional and geometric attacks.