A Reversible Digital Watermarking Algorithm Based on Medical Images via Threshold Segmentation and Contrast Enhancement.

With the popularization of the Internet, the information is more convenient to acquire and transmit via network. But the privacy protection under the digital environment should be paid more and more attention. Now, in the paperless hospital, the reversible digital watermarking technology is adopted to embed the electronic cases into medical images of patients as watermark form, which can ensure the privacy protection. However, the existing reversible digital watermarking algorithm usually pursues the peak signal-to-noise ratio, which makes it difficult to consider the embedding capacity and image quality. To solve these problems, a reversible digital watermarking algorithm based on threshold segmentation for image contrast enhancement is proposed. In the proposed model, the Ostu threshold segmentation is applied to the gradient amplitude of the image to obtain the pixel threshold, and the bidirectional scanning is used to determine the foreground area. Meanwhile, the square column displacement between peak-pixel and zero-pixel value is used to avoid pixel overflow. The redundant information is added in the process of embedding to ensure the accuracy of information extraction. The experimental results show that the algorithm can keep a higher embedding capacity and make the embedded image with a higher visual quality.