Watermarking of radiographic images based on quantum noise modulation

Abstract

In this paper, we propose a new watermarking scheme for digital radiographic images which takes advantage of the quantum noise. This later is inherent to their acquisition process and is also the dominant noise. Once this noise extracted from the image, we modulate it for message embedding with Quantization Index Modulation under the constraint the watermarked noise respects the original noise distribution. By doing so, it is expected that the image diagnosis value is preserved. Experiments conducted on a set of 75 radiology images of different anatomical structures of the body show that the image distortion, evaluated through several image quality metrics, is very low (PSNR∼77.24dB) while offering a relatively important capacity (capacity∼0.15bpp) that can support different watermarking based security objectives.