摘要 |
Video containing frames of blocks of pixels is transmitted from encoder to decoder. The encoder extracts motion vectors from a frame 1(t-1) preceding the frame 1(t) being encoded, and creating an irregular grid of cells, the cell sizes based on motion information, e.g. complexity in the frame, at a respective position. This gives a motion vector field of an irregular grid of differently-sized cells, each cell associated with a motion vector. The motion vectors are transmitted to the decoder as auxiliary information along with usual motion prediction information (i.e. motion vector field and block residuals) of the frames (at least frames 1(t-2), 1(t-1) and 1(t)). The decoder receives the motion prediction information, with missing slice, and auxiliary information, and rebuilds the irregular grid for frame 1(t) based on the frame 1(t-1) as the encoder did and fills the cells with the motion vectors from the auxiliary information, thus recreating an estimated motion vector field for the current frame 1(t) for subsequent error concealment/decoding/displaying. An encoded video bitstream with minimum auxiliary (i.e. error concealment) information is thereby sent, thus enabling frame reconstruction of the video bitstream if a slice of the frame is lost. |