International Journal of Computer Applications (0975 – 8887) Volume 57– No.11, November 2012 1 Enhanced Digital Image Watermarking Scheme based on DWT and SVD Satendra Kumar Asst. Prof., Dept. of CSE Seemant Institute of Technology, Pithoragarh Jaydeep Kishore Asst. Prof., Dept. of CSE Seemant Institute of Technology, Pithoragarh Nitin Arora Asst. Prof., Dept. of CSE Women’s Institute of Technology, Dehradun ABSTRACT Digital watermarking is a technique of embedding some information (usually hidden copyright notices) into an image. Number of applications has been found in various fields like copyright protection, content authentication, document annotation, medical imaging. So, an enhanced semi-blind, hybrid digital image watermarking scheme based on discrete wavelet transform (DWT) and singular value decomposition (SVD) is our proposed approach in this paper. To increase and control the strength of the watermark, we used a scale factor. our proposed approach, the watermark is not embedded directly on the wavelet coefficients but rather than watermark wavelet coefficient are inserted on singular values of the cover images with modifying wavelet transform (DWT) HL1 sub band. Experimental results clearly show that this proposed scheme is quite resilient to various image processing attacks. Keywords Watermarking, Wavelet transform, Scale factor (SF), singular value decomposition (SVD), Peak signal to noise ratio (PSNR), Normalized cross correlation (NCC). 1. INTRODUCTION Internet has made it extremely easy for anyone to share any kind of data with or without the consent of its owner; to curb this menace of unauthorized data transmission digital watermarking [2], [3] came out as a solution. In digital image watermarking we embed a hidden copyright notice onto the cover image. Later it could be extracted to claim the ownership of the image. It is also being used for content authentication, document annotation, biomedical image processing [4], [5], secret communication [1]. A good watermark should be robust and imperceptible. Robustness refers to resilience of the watermark towards common image processing attacks as well as intentional attacks. In particular, watermark should still be detectable or extractable even after common signal processing operations (such as geometric transformations, cropping, compression, filtering, resampling, scaling, rotation etc.) have been applied to the watermarked image. Imperceptibility means that it should not affect the visual quality of the cover image due to its presence. Robustness and imperceptibility of the watermark have a trade-off between them. It is desirable for a watermarking scheme to achieve a balance between both these properties. Watermarking can be categorized into spatial domain or transform domain depending upon how the watermark is being embedded. In spatial domain the watermark is directly embedded onto the pixel level itself. It has the advantage that it is fast and easy to implement but the embedded watermark gets distorted even in the slightest introduction of noise or in case of any tampering. In transform domain, watermark is added onto the coefficients of the cover image obtained after the transformation e.g. discrete cosine transform (DCT), discrete wavelet transform (DWT), and singular value decomposition (SVD). Transform domain techniques are usually preferred over spatial domain techniques because of the fact that they are more resilient to various image processing attacks. One of the important properties [6] of the SVD is that singular values of a matrix do not exhibit prominent change when slight variation in the matrix elements is being carried out. This property encourages watermark insertion onto the singular values of the matrix. In this paper we propose a hybrid digital image watermarking technique combining both these transformation methods. Experimental results have also been provided to prove extreme robustness of this proposed scheme. A digital watermark is an unnoticeable signal added to digital data, known as cover work, which can possibly be identified at a later stage for ownership proof. Ideal properties of a digital watermark: 1. A digital watermark should be imperceptible, meaning that it should be perceptually invisible to prevent obstruction of the original image. 2. Watermarks should be robust to filtering, additive noise, compression, cropping and other forms of image manipulations. Watermarks can be categorized into non blind, semi-Blind and blind schemes on basis of the requirements for watermark extraction [7]. This paper, proposed embedding scheme uses a two level Discrete Wavelet decomposition is presented. One of the main contributions of the proposed technique is in decomposing the host image. Second main contribution of proposed scheme the watermark is not embedded directly in SVD wavelet coefficient rather than the value of inserted Watermark Wavelet coefficient are inserted in SVD 2 nd Level Wavelet coefficient elements of the Cover image. Primarily the proposed technique aims to improve the robustness of the