Real-time Protection of Scalable Video Content

This work is focused on protection of state of the art video bitstreams and their scalable extensions. For copyright protection and meta-data embedding, we opt for watermark embedding with very controlled compromise on PSNR and bitrate. The secret lies in the fact that the reconstruction loop has taken into account to avoid the mismatch on the encoder and decoder. Spread spectrum watermarking was then implemented on QTCs inside the reconstruction loop. It is, later on, used to embed Tardos fingerprinting code to avoid traitor tracing. For restriction of access only to authorized users, selective encryption of H.264/AVC and AVS video content is proposed. For this purpose, the entropy coding module has been modified to transform a video codec into a crypto-compression module, which can perform the encryption and compression at the same time. The encrypted bitstreams are fully format compliant with exactly the same bitrate. In addition, the extra processing power required for encoding and decoding encrypted bitstreams is very negligible. Algorithms for independent protection of enhancement layers in a scalable video have also been proposed in this work.

Autorentext

Dr. Zafar Shahid received his PhD from University of Montpellier II, MS from INSA Lyon France and his B.Sc. in Electrical Engineering from UET Lahore (Pakistan). His research interests include image/video processing and embedded systems.