Hi all, Recently, I read some articles about MCTF structure which was used in the extension part of H.264/AVC standard. This scalable codec use MCTF to achieve the temporal scalability and claim that this structure will not have the drift error like conventional hybrid coder do. I wonder why this structure can make such a claim because the error in the low-pass picture or lower layer high-pass picture will also propagate in the reconstruction of the final result. I also want to know that how could this structure generate the residual signal only by using the original frame instead of the reconstructed one. As we know in the hybrid coder, transmitting the residual of two original frame will make the encoder and decoder in the mismatch situation and that will crash the codec. I hop here are some one who can help me and let me thank in advance. Regards, Bacchus Lee |
drift error in Motion Compensated Temporal Filtering (MCTF) structure used in Scalable Video Coding
Started by ●March 3, 2005