A Digital Watermarking Approach using SMQT, OTSU, DWT and IDWT (Well Researched Paper with Accurate Citations and References)

In this paper, a new digital watermarking model is proposed for the medical images. An improved SMQT is used for image enhancement and the image is being segmented using OTSU thresholding. Discrete Wavelet Transform (DWT) and Inverse DWT are used to embed and extract the watermark on the host image. The goal of our scheme is to make the watermarking more robust against attacks and secure the image from privacy threats. In the experimental evaluation, we tested our proposed approach with various attacks on the watermarked image and then host image is extracted and matched with the original image. Experimental results demonstrate that our proposed model outperforms the existing state-of-art models by exhibiting higher PSNR, lower MSE and improved CC values. I. INTRODUCTION Nowadays, medical images are transmitted over electronic networks for improved health care as well as for clinical interpretation. Digital medical imaging facilities have become so reliable and less costly that film-based imaging technology has transferred to filmless technology for producing digital images. For the medical image watermarking scheme, recognizing image source and proof of identity of the correct patient are needed [1]. These features make the watermarking of medical images more challenging and sophisticated. Therefore, it is important to find and develop new watermarking models which can fulfill these requirements since there is no single approach which can be termed as best. A number of researchers worked on developing digital watermark on medical images. Spatial domain and transform domain are two commonly used methods in watermarking and transform domain methods are more prevalent than spatial domain because watermarked image becomes more robust when this domain is used [2, 3]. One of the frequently used algorithms called the histogram equalization method. However, phenomenon of over saturation, noise, and distortion frequently appears. A mathematical scheme of histogram equalization optimization was