Stationary Wavelet-Based Image Watermarking for E-Healthcare Applications

Abstract

Watermarking is a technique for protecting multimedia data, mainly images, from malicious attacks by adding a signature into these images. Traditional watermarking techniques, unfortunately, have drawbacks when used on sensitive images like those used in medicine. A reliable and blind watermarking method is suggested in this work to protect medical picture transferred in telemedicine. Medical picture marking enables accurate patient identification, prevents scan confusion, and minimizes the risk of diagnostic mistakes that could have negative consequences. In this method, three distinct transforms are used to acquire the frequency content of the image. The low frequency subbands are subsequently subjected to Schur decomposition. In order to incorporate the watermark bits, the resultant upper triangular matrix values are modified. The proposed approaches effectively retain a considerable quality of watermarked images and are impressively resistant against numerous conventional attacks, according to imperceptibility and robustness experimental results. The average Peak signal-to-noise ratio (PSNR) which was obtained was 44.90 dB, demonstrating that the integration procedure produces very low distortion to the original image. The observed results for robustness demonstrate that the watermark is resilient to the most of the attacks performed in watermarking, with a normalized cross correlation rate higher than 0.9.