Healthcare Image Watermarking Scheme Based on Human Visual Model and Back-Propagation Network

In this paper proposed a new healthcare image watermarking scheme based on human visual model and back-propagation network technique. The human visual model which is characterized by properties like luminance, frequency, and texture sensitivities was utilized to generate the suitable strength of an embedded watermark. In addition, the back-propagation network technique was employed to obtain the local characteristic of an image. Experimental results showed that the embedded watermark proposed healthcare image that could survive several kinds of image-processing attacks and the JPEG lossy compression. The proposed image watermarking scheme can be used on the Internet to reduce difficulties that medical personnel may encounter.