Research On Video Watermarking Technology For Streaming

In recent years, with the rapid development of various streaming media industries including video-on-demand, Internet TV and video sharing site, the problem of copyright protection for streaming video becomes increasingly serious. As an efficient protection method, video watermark, when it is used in streaming, not only needs to meet the basic requirement of traditional video watermark like imperceptibility, but also should be adapted to the specific application environment. For example, in order to identify rebellious users who illegally copy videos in the video-on-demand application, the user identity information needs to be quickly embedded into video with real-time. Meanwhile, in the social networking site, large amount of videos are uploaded every day. So in order to protect the copyright of these videos, watermark embedding should not affect the normal operation of the website and the watermark can be correctly extracted after various processing as well. Thus, according to the streaming videos of the video-on-demand and social networking site, two video watermark algorithms are respectively proposed in this thesis with the following works:(1) A real-time video watermark algorithm based on the motion vector is put forward to track piracy of video in video-on-demand application. Firstly, the concept of watermark embedding margin of frame (WEMF) is proposed to shorten the watermark embedding time. By calculating the value of WEMF for each B frame, which is related to its frame bits, suitable B-frames can be selected to embed watermark without decoding it completely. Moreover, with the characteristic of variable macro block size in H.264coding standard, a simplified formula is presented to quickly calculate the just noticeable difference (JND) value of the macro block, and then the macro blocks with larger JND value are selected to embed watermark into their motion vectors, which also provides good invisibility. Experimental results show that the proposed algorithm can find suitable embedding positions and embed watermark quickly with good invisibility and less increase of the bit-rate of video. The algorithm can be used for real-time watermark embedding in the video-on-demand application.(2) A video watermark algorithm based on video region DCT coefficients is proposed aiming at protecting the legitimate right for the video producer in the social networking site. Firstly, the video frame is divided into regions before watermark embedding.The algorithm selects the appropriate video region DCT coefficient to embed watermark by calculating the energy of each region.The watermark is embedded dispersed to multiple macro blocks, so the robustness of the watermark is enhanced. Besides, the video region DCT coefficients are modified by dynamic quantization step size, which improves the imperceptibility of watermark. Watermark is embedded while video is transcoding, which avoids secondary decoding and compression of video. What’s more, the watermark is embedded by the video producer, which does not affect the normal operation of the website. Experimental results show that the watermark can be correctly extracted after common operation, such as scaling, weight and transrating, at the same time, the video quality degrades little. The algorithm can be used for video copyright protection in the social networking site.