The recent developments of enormous computer networks have invoked insecurities related to copyright theft of digital media. To be precise, the virtual sharing of medical images over networks with a novel desire of improved medical diagnosis has led to the tampering of sensitive patient identity information. In this chapter, the authors have exemplified the need of watermarking with fragile medical image watermarking using saliency and phase congruency. Initially, the saliency and phase congruency methodologies are applied on the original medical image to highlight the object features. Based on the feature map, a mask is generated which segregates the area of interest from the portions containing visual medical information. An encrypted text, containing identity of the patient, is embedded into the area of interest of the image. The results of imperceptibility and fragility criteria are satisfactory towards the implementation of a fragile watermark as the extracted watermark is found to be corrupted upon unfaithful image processing modifications.
A feature adaptive image watermarking framework based on Phase Congruency and Symmetric Key Cryptography
