Adaptive-Rate Image Watermarking based on Spread Spectrum Communication Technique

As the Internet becomes more and more populous, people concern more about the copyright protection issue for digital data such as images and audio. Digital watermarking technique can hide data in images or audio to indicate the data owner or recipient. Therefore, it can protect the copyright. There were a lot of papers discussing how to embed watermark into images in recent years. However, few papers analyzed the watermark techniques in a theoretical point of view. In this paper, we rst interpret the watermarking problem as a digital communication problem. There are three main criteria concerning the performance of a watermarking techniquecapacity, imperceptibility, and robustness. We then show the trade-o 's between these three criteria adopting concepts from the digital communication theory. After the analysis has been established, an adaptive coding-rate watermarking scheme based on spread spectrum communication technique is proposed. Consider the original image as transmission channels, capacity is interpreted as channel capacity of the image, imperceptibility is the SNR with watermark as signal and original image and imageprocessing attacks as noise, and probability of watermark detection error demonstrates the robustness of the watermarking technique. An analysis shows that we cannot increase the capacity without making the watermark more obtrusive in viewing. In addition, the watermarking system is more robust when the watermark signal power rises. Unlike the general digital communication system, where the channel characteristics is known or estimated, an image channel is hard to de ne because it can be transformed to many di erent domains, such as DCT domain, Wavelet-Transform domain, etc. Therefore, an image channel has di erent capacity and probability of error respect to di erent representations. In this paper, we develop our system on two most commonly used domains, spatial and DCT domain. An image has di erent capacities in di erent spatial or frequency components. We use di erent coding rates for watermarking signals in di erent components to achieve the channel capacity we have calculated in the previous chapter. The experiment results show more bits embedded and lower probability of error than the conventional xed-rate coding. Key-Words: watermark, adaptive rate coding, data hiding, stegonagraphy, copyright protection, human visual system (HVS), channel capacity, signal to noise ratio (SNR), probability of error