Digital Watermarking for Medical Images using Biorthogonal Wavelet Filters and Transformed Watermark Embedding
Mahavir Dhoka, Jitendra Kadam
Abstract
In recent years, data protection is one of important aspect because of cases like piracy, copyright and ownership issues. Digital watermark seems to be solution to the problem. In our paper, we proposed a method of invisible watermark to medical image using Biorthogonal wavelet filter and transformed domain watermark embedding. Generally watermark is embedded in original image either directly or in transform of original image. In our method, we transformed both original image and watermark using discrete wavelet transform and Biorthogonal wavelet filter coefficients. We specifically follow this method considering the case of medical images. As medical images are low contrast images and are taken at high precision with special equipment, so it is not expected that any one will claim for its ownership. Hence we tried to develop a method which will recover watermark through medical image considering any kind of attacks. We tested our method on multiple medical images and listed down results and comparison is made on basis of PSNR and normalised correlation (NC). From the values of NC, we concluded that our method gives better recovery of watermark from watermarked image.
Keyword
Invisible Watermark, medical images, DWT, BWT, PSNR, NC, Watermark Attacks.