A Robust Dual Watermarking Technique for Medical Images in E-Healthcare Records

This paper proposes a prudent implementation of robust pipeline in the domain of watermark based secure medical image transmission. The pipeline comprises of three stages, involving insertion of first watermark in the first stage and the second watermark in the second stage. Insertion of the watermark invokes a multi transform algorithm implemented in two steps. The first step is the application of Two-Way Arnold Transformation (TWAT) which scrambles the pixel locations in the source image, based on the number of cycles passed as one of the arguments to it, leading to a reduction in spatial correlation. The second step is 2-D Wavelet Transformation (WT) applied to the scrambled pixels that exploit the reduced correlation feature to generate the large values of coefficients. These large coefficient values can efficiently accommodate the wavelets coefficients of the inserted watermark. This step is a novel contribution compared to the existing models. Subsequently, the inverse WT is applied which concludes the first stage of the pipeline. The second stage of the pipeline also implements the same sequence of steps as that of the first stage but with the second watermark. In the final stage of the pipeline, the inverse TWAT is applied twice the number of times to the forward transformations applied in the first two stages. The entire pipeline is termed as Multi Transform based Robust Dual Water Marking (MTRDW). The pipeline is implemented on benchmark images with patient information and hospital logo as watermarks. The claimed features of the model are presented through computing the quality metrics such as SSIM, NCC and PSNR from the experiments conducted.