ROBUST WATERMARKING OF GRAYSCALE IMAGES USING FEATURE EXTRACTION AND DWT

The rapid advancement and extensive growth in communication technology is creating a critical need to develop novel methods and techniques to protect copyright, ownership, authentication and content integrity of digital media. An efficient solution to this kind of problem is digital watermarking of multimedia. A watermark refers to a visible or invisible digital data to be embedded in a cover data so as to provide copyright protection at time when need arises. Numerous watermarking and steganographic methods have been presented in the literature and several watermarking software packages have been developed in past decades. Still robustness and security of watermark against a variety of attacks is one of the most important issues to be solved. In this paper, grayscale 8 bit images are considered for watermark embedding. However, techniques are proposed so that the method is directly applicable for watermarking of color images. Discrete wavelet transform is used to provide robustness and Noise Value Function (NVF) and Arnold Scrambling is used to provide security. Watermark generation is done using LL band of NVF function of original image. The watermark is then subjected to Arnold Transform to introduce randomness, thus enhancing security and improving PSNR. Embedding is done in low frequency LL band of original image. Inverse Discrete wavelet Transform is then computed to obtain the watermarked image. The results are evaluated by applying number of attacks on the watermarked image.