Method and apparatus for encoding/decoding image using variable-size macroblocks

摘要

The video encoding/decoding apparatus includes a video encoder for encoding macroblocks by each of prediction basic unit candidates, determining prediction basic units based on encoding costs for each of the prediction basic unit candidates, and generating a bitstream including image data encoded in determined prediction basic units, information on a prediction type for each of the determined prediction basic units, and prediction data depending on the prediction type; and a video decoder for extracting information on the prediction type for each of the prediction basic units, prediction data depending on the prediction type, and encoded image data from the bitstream, and generating a reconstructed image by decoding the encoded image data depending on the prediction type for each of the prediction basic units.

主权项

1. A video decoding method using a tree structure, the method comprising:
decoding, from a bitstream, a partition information for indicating a subblock to be decoded within a macroblock, wherein the partition information includes a partition flag indicating whether each block, starting from the macroblock corresponding to an uppermost layer of the tree structure, is divided into four blocks of a lower layer, and a block which is not further divided in the tree structure is identified as the subblock within the macroblock; decoding prediction type information indicating whether the identified subblock is intra predicted or inter-predicted, and prediction mode information required for an intra prediction or an inter prediction, from the bitstream; and reconstructing the identified subblock by using the prediction type information and the prediction mode information, wherein, when a block of a first layer in the tree structure is divided into four blocks of a second layer which is a lower layer of the first layer, the partition flag for each of the four blocks of the second layer is decoded in a predetermined order, and wherein, when a block among the four blocks of the second layer is further divided into four blocks of a third layer which is a lower layer of the second layer, the partition flag for each of the four blocks of the third layer is decoded in the predetermined order, prior to decoding the partition flag for a next block of the second layer in the predetermined order, wherein the predetermined order is an order of an upper left block, an upper right block, a lower left block and a lower right block, wherein the partition flag for a block having a size identical to a minimum block size is not decoded from the bitstream.