Method and apparatus for processing a video signal
An apparatus for processing a video signal and method thereof are disclosed. The present invention includes determining an intra reference block in a current view layer by using an intra motion information of an inter-reference unit in a reference view layer, the inter-reference unit being partially decoded and decoding the current block in the current view layer by using the intra-reference block. Accordingly, in coding a video signal, pictures failing to be substantially referred to for reconstruction of a current picture can be omitted. According to another aspect of the present invention, residual information having high redundancy is omitted to raise a compression ratio.
1. A method of decoding a multi-view video signal, comprising:
obtaining anchor picture flag information of a current picture in a current view from the multi-view video signal, the anchor picture flag information indicating whether the current picture is an anchor picture or a non-anchor picture;
obtaining motion skip information of a current block in the current picture when the anchor picture flag information indicates the current picture is the non-anchor picture;
generating motion information of the current block by using motion information of a corresponding block in a neighboring view when the current block is in a motion skip mode according to the motion skip information;
determining a reference block in the current view based on the motion information of the current block; and
decoding the current block in the current view by using the reference block,
wherein the generating including:
determining the corresponding block in the neighboring view based on view dependency and a global motion vector of the current picture, the view dependency being determined based on reference number information and view identification information for a plurality of views, the global motion vector of the current picture being derived from a global motion vector of the anchor picture; and
obtaining the motion information from the determined corresponding block.
2. The method of claim 1 , wherein the reference block exists on other pictures in the current view except the current picture having the current block located thereon.
3. The method of claim 1 , wherein the motion information comprises motion vector, block type information, and reference index.
4. The method of claim 1 , wherein the corresponding block in the neighboring view is at a same temporal location as the current block in the current view.
5. The method of claim 1 , wherein the reference number information indicates a number of reference views and view identification information is a view identifier of each view.
6. The method of claim 1 , wherein the multi-view video signal is received as a broadcast signal.
7. The method of claim 1 , wherein the multi-view video signal is received on a digital medium.
8. A non-transitory computer-readable-medium comprising a program recorded therein to execute the method of claim 1 .
9. An apparatus for decoding a multi-view video signal, comprising:
a decoding apparatus configured to,
obtain anchor picture flag information of a current picture in a current view from the multi-view video signal, the anchor picture flag information indicating whether the current picture is an anchor picture or a non-anchor picture,
obtain motion skip information of a current block in the current picture when the anchor picture flag information indicates the current picture is the non-anchor picture,
determine a corresponding block in a neighboring view based on view dependency and a global motion vector of the current picture when the current block is in a motion skip mode according to the motion skip information,
obtain motion information from the determined corresponding block,
generate motion information of the current block by using the motion information of the corresponding block in the neighboring view,
determine a reference block in the current view based on the motion information of the current block, and
decode the current block in the current view by using the reference block,
wherein the view dependency is determined based on reference number information and view identification information for a plurality of views, and
wherein the global motion vector of the current picture is derived from a global motion vector of the anchor picture.