| 摘要 |
The present invention relates to a decoder for a multi-layer video, an inter-layer prediction device for a multi-layer video and an inter-layer prediction method for a multi-layer video. The decoder for a multi-layer video according to the present invention comprises an inter-layer division structure prediction part for determining a division structure of an improvement layer by referring to the division information of a reference layer and an inter/intra prediction part for determining a prediction mode based on the division structure of the improvement layer. The inter-layer prediction method for a multi-layer video comprises: a division structure use decision step for determining the use of reference division information when the video of the improvement layer is divided based on the division information of the reference layer; a corresponding position structure selection step for determining a division unit of the reference layer of a position corresponding to the division unit of the improvement layer; and a division structure prediction level selection step for selecting the reference of the division information of the coding unit and converting unit of the reference layer and transmitting the same to a dequantization part. [Reference numerals] (101,201) Entropy decoding part; (102,202) Dequantization part; (103,203) Inversion part; (104,204) Motion compensation part; (105,205) Intra prediction part; (106,206) Deblocking filter part; (107,207) Sample adaptable offset part; (108,208) Restoration picture buffer; (300) Demultiplexer; (400) Inter-layer division structure prediction part; (AA) Bit stream |
