This paper proposes a number of video transcoding techniques for the purpose of adding error resiliency. The proposed solutions make use of distributed video coding technologies which were originally reported in the literature for distributing coding complexity between the encoder and the decoder. Three transcoding solutions are proposed: frequency-domain, time-domain and compressed-domain transcoding, and various decoding architectures are investigated. Error resiliency at the decoder is addressed as a post-process, an integrated process and a pre-process. It is shown that implementing error resiliency as a decoder pre-process and combining this with compressed-domain transcoding removes the need to interfere with the functionality of existing and compliant decoders whilst minimising complexity. The proposed solutions serve as a framework for boosting the error resiliency of pre-encoded video and can be applied to MPEG-2, and H.264/AVC coded streams.

This paper presents an extension to video error resiliency based on distributed source coding. The paper proposes a frequency domain transcoder and a compressed domain transcoder for boosting the error resiliency of pre-encoded video using distributed source coding techniques. In both transcoders it is proposed to identify regions of interest in the pre-encoded video. In the compressed domain transcoder, the regions of interests are retained as is, whilst in the frequency domain transcoder, such regions are reproduced at a lower fidelity by requantization or ideal low pass filtering. Channel coding follows, and the parity bits are transmitted in an error resiliency supplementary bitstream. At the destination, the paper proposes a pre-decoding solution that regenerates the low fidelity ROIs and corrects the received bit stream using the aforementioned supplementary stream. The corrected bitstream is then repacked and sent to the destination decoder. In comparison with traditional FEC techniques, experimental results show robust resiliency against packet losses up to a loss ratio of 15%. IntroductionVideo compression with distributed source coding is a paradigm in which the computational complexity is distributed between the encoder and the decoder [1]. Distributed video coding found other applications in error resiliency as well. For instance, the work in [2] is based on coding raw video using AVC whilst employing distributed source coding as an error resilience mechanism. Regions Of Interest (ROI) are selected from the partially reconstructed video using Flexible Macroblock Ordering (FMO). Redundant slices are then used to re-encode the ROI using the coding modes of the original video albeit at a lower quality through coarser quantization. The coarse representation of the video is then channel coded and the parity bits alongside the new quantization parameters and region boundaries comprise the Error Resiliency (ER) supplementary bitstream. The decoder on the other hand with the guide of the received supplementary bitstream reconstructs the coarse video representation and applies channel decoding and correction to it. If needed this coarse representation replaces the part of the bitstream that was received in error. Clearly, a motion compensation loop-mismatch is anticipated here, however such a mismatch has a less severe visual effect than that caused by the bitstream errors. This paper proposes to apply error resiliency based on distributed source coding to pre-coded video using video transcoding techniques. We also propose a complementary pre-decoding approach to ensure the compliancy of the destination decoders. This can be applied to both AVC and MPEG-2 pre-encoded video. The solution is also important to boost the error resiliency of broadcast video services such as

No abstract