Video Watermarking Based on Neural Networks

Abstract

In this paper, we propose a novel digital video watermarking scheme based on multi resolution motion estimation and artificial neural network. A multi resolution motion estimation algorithm is adopted to preferentially allocate the watermark to coefficients containing motion. In addition, embedding and extraction of the watermark are based on the relationship between a wavelet coefficient and its neighbor's. A neural network is given to memorize the relationships between coefficients in a 3x3 block of the image. Experimental results show that embedding watermark where picture content is moving is less perceptible. Further, it shows that the proposed scheme is robust against common video processing attacks. INTRODUCTIONThe sudden increase in watermarking interest is most likely due to the increase in concern over copyright protection of content. With the rapid growth of the Internet and the multimedia systems in distributed environments, digital data owners can easy transfer their multimedia documents across the Internet. However, current technology does not protect their copyrights properly. This leads to wide interest of multimedia security and multimedia copyright protection and it has become a great concern to copyrights owners in recent years. In the early days, encryption and control access techniques were used to protect the ownership of media. Digital watermarking, the art of hiding information in a robust and invisible manner, has been investigated as a complementary technology [1]. There are some desirable characteristics of effective watermarking techniques, including imperceptibility, robustness, and security [2]. Imperceptibility is the degree of perceptual similarity between the original and watermarked signals. It is desirable to embed the watermark in a discreet, unobtrusive manner so that the watermark is imperceptible under casual observation. Robustness is the resilience of the watermark against manipulations such as lossy compression, linear and nonlinear filtering, scaling, and cropping (Stirmark [3], dewatermarking attack [4]).Security is the ability of the watermark to resist hostile attacks. Attacks are not limited to removal of the watermark, but include watermark estimation or forgery, collusion, and ambiguity attacks.Obviously, it is desirable to have an imperceptible, robust, and secure watermarking technique.Digital watermarking has recently been extended from still images to video. Many algorithms have already been proposed in the literature. It goes from the simple adaptation of watermarking algorithm initially designed for still images [5] to the definition of specific video watermarking scheme [6]. Open paths still remain in video watermarking. This technology is indeed in its infancy and is far from being as mature as for still images.In this paper, we propose a novel approach to neural network watermarking for uncompressed video in the wavelet domain. The watermark is embedded multiple times into the host data. Before actually embedding the watermark information, it take...

A hybrid approach for Content-Based Image Retrieval based on Fast Beta Wavelet network and fuzzy decision support system

“…In other side, Philippe and Matthieu [50] introduced in their paper the most known active learning methods for image retrieval such as Bayes classification, k-Nearest Neighbors [51], neural networks [52,53], wavelet network [54,55], lattice trees [56][57][58], Gaussian mixtures and support vector machines. Ekta and Hardeep [59] proposed the use of bayesian algorithm, as a supervised learning and a statistical method for classification, by reducing the noise from images.…”

An Analysis on Digital Watermarking in Videos based on Frequency Domain Techniques

“…Experiments shows that it is robust to video watermarking attacks like lossy compression, frame averaging, noise addition and frame dropping. The introduction of classification system based on Artificial Neural Network (ANN) by M. El'arbi et al [31] has done a good job in selecting the transformation method which is best for embedding process. Experiments shows that the selected transformation technique increases the quality of video as well as show robustness against geometric attacks.…”

Computational Intelligence in Multimedia Processing: Foundation and Trends

“…Maher et al [31] proposed a novel digital video watermarking scheme based on multi resolution motion estimation and artificial neural network. A multi resolution motion estimation algorithm is adopted to preferentially allocate the watermark to coefficients containing motion.…”