US 7,321,625 B2
Wavelet based multiresolution video representation with spatially scalable motion vectors
Huipin Zhang, Milpitas, Calif. (US); and Frank Jan Bossen, San Jose, Calif. (US)
Assigned to NTT DoCoMo, Inc., Tokyo (Japan)
Filed on Dec. 13, 2002, as Appl. No. 10/318,802.
Prior Publication US 2004/0114689 A1, Jun. 17, 2004
Int. Cl. H04B 1/66 (2006.01); H04N 7/12 (2006.01); H04N 11/02 (2006.01); H04N 11/06 (2006.01)
U.S. Cl. 375—240.11 23 Claims
OG exemplary drawing
 
1. A method for determining multiresolution video representations for scalable video coding, the method comprising:
performing subband motion compensated temporal filtering (MCTF) on input video signals, including performing a spatial decomposition on the input video signals, wherein motion vectors are available for subbands of the spatial decomposition; and
obtaining a multiresolution representation of the filtered input video signals along both spatial directions and temporal direction,
wherein spatial scalability is supported by video coders that use the multiresolution video representations, and, when combined with temporal scalability, is limitedly supported by video coders, depending on the representation that is used by the video coders, and further wherein combined spatial scalability and temporal scalability supported by a representation can be characterized by a spatial-temporal scalability (ST) table.