Dutta, Tanima2015-09-222023-10-202015-09-222023-10-202014ROLL NO. 10610106http://172.17.1.107:4000/handle/123456789/500Supervisor: Arijit SurIn recent years, digital watermarking is being regarded as a promising solution in order to mitigate the increasing threats of video piracy and for ensuring video ownership and content authentication. Digital videos are generally stored and transmitted in a compressed format, which encourage research in designing video watermarking algorithm in compressed domain to avoid the complete decoding and re-encoding of a compressed video. H.264/AVC is a recent and an e cient video compression standard which provides higher compression than the MPEG-2 and MPEG-4, maintaining acceptable perceptual quality. The performance of video watermarking is often evaluated with parameters such as robustness, imperceptibility, security, transparency, payload, increase in video bit rate, blindness, etc. Most of these parameters are interdependent, often con icting and are chosen based on the application. In this dissertation, di erent issues in compressed domain watermarking for H.264/AVC encoded video have been addressed. In this work, a few video frames (I-frame and P-frame) based watermarking algorithms have been designed and di erent challenges associated with them have been solved. In the rst part the dissertation, a P-frame based robust and blind watermarking method is proposed. The proposed method not only provides controlled increase in video bit rate and better perceptual quality of watermarked video, but can also withstand frame dropping, averaging and swapping (FDAS) attack through a robust watermarking method that uses repeat accumulate code and erasure channel.enCOMPUTER SCIENCE AND ENGINEERINGThesis