An Attack Resistant Data Hiding Technique for Medical Images

Abstract

Medical image processing has been gaining widespread significance and attention with the advent of new technologies to increase speed and lower the cost and time. At the same time, there has been no compromise in the loss of quality on the processed medical image as they remain to be very sensitive in the sense that even the slightest degradation as a result of the processing would lead to fatal consequences. Numerous techniques have been put forward over the time to make the image under study resilient towards a wide range of attacks especially when some vital information is concealed or embedded inside the host medical image. A robust technique is proposed which is ablt to provide the necessary robustness to the host image. An invisible non blind multi resolution watermarking scheme for medical images using Contourlet transform (CT) is proposed in this paper. The directional features of the Contourlet transform have been utilized to provide choice of embedding location in a hybrid combination with the discrete cosine transform (DCT) to provide the required robustness. The embedded image along with its payload has been tested with a wide range of intentional and unintentional attacks simulating the real time attacks and the results justify its superiority over the existing techniques in terms of robustness. The evaluation has been done using standard metrics like Peak signal to noise ratio (PSNR), Correlation coefficient (CC). Since the proposed method has good resistance to many of the attacks and good perceptibility it is suitable for data hiding in medical image.