Method and apparatus for decoding/encoding a video signal
The present invention provides a method of decoding a video signal. The method includes the steps of obtaining view information of a picture from the video signal and generating information for reference picture management using the view information.
1. A method for decoding a multi-view video data in a multi-view video stream with a decoding apparatus, comprising:
receiving, with the decoding apparatus, the multi-view video stream, the multi-view video stream including a random access picture, the random access picture being a coded picture in which all slices reference only slices having a same time as the random access picture;
obtaining, with the decoding apparatus, a random access flag for inter-view prediction from the multi-view video stream, the random access flag indicating whether a type of picture is the random access picture;
obtaining, with the decoding apparatus, initialization information of a reference picture list, the initialization information representing view dependency relationships between a plurality of views, the initialization information including view number information and view identification information;
initializing, with the decoding apparatus, the reference picture list based on the random access flag, the initializing including ordering reference pictures for the random access picture in a decoded picture buffer, the ordering including assigning an inter-view reference index to each reference picture, the inter-view reference index identifying a reference picture in the reference picture list used for inter-view prediction of a block to be decoded in the random access picture, the assigning the inter-view reference index being performed using the view number information and the view identification information;
obtaining, with the decoding apparatus, modification information for the initialized reference picture list from the multi-view video stream, the modification information representing how to change the inter-view reference index assigned to the reference picture in the initialized reference picture list;
modifying, with the decoding apparatus, the initialized reference picture list for inter-view prediction, the modifying including reordering the reference picture by changing the inter-view reference index assigned to the reference picture based on the modification information;
determining, with the decoding apparatus, a prediction value of a current macroblock based on the modified reference picture list; and
decoding, with the decoding apparatus, the current macroblock using the prediction value,
wherein the initialization information is obtained from an extension area of a sequence header in the multi-view video stream.
2. The method of claim 1 , wherein the view number information indicates a number of reference views of the random access picture, and the view identification information provides a view identifier of each reference view for the random access picture.
3. The method of claim 1 , wherein the multi-view video data includes video data of a base view independent of other views, the base view being a view decoded without using inter-view prediction.
4. The method of claim 1 , wherein the inter-view reference index is changed by performing a subtraction operation or an addition operation between a view information difference and a predicted inter-view reference index,
wherein the view information difference means a difference between a inter-view reference index of a current reference picture in the initialized reference picture list and the predicted inter-view reference index, and
wherein the predicted inter-view reference index indicates a inter-view reference index of the reference picture assigned right before.
5. The method of claim 1 , wherein the modification information is obtained from a slice header.
6. The method of claim 1 , wherein the modifying step shifts other remaining pictures in the initialized reference picture list to later positions in the initialized reference picture list.
7. An apparatus for decoding a multi-view video data in a multi-view video stream, comprising:
a parsing unit configured to receive the multi-view video stream, the multi-view video stream including a random access picture, the random access picture being a coded picture in which all slices reference only slices having a same time as the random access picture, the parsing unit configured to obtain a random access flag for inter-view prediction from the multi-view video stream, the random access flag indicating whether a type of picture is the random access picture, and the parsing unit configured to obtain initialization information of a reference picture list, the initialization information representing view dependency relationships between a plurality of views, the initialization information including view number information and view identification information;
a decoded picture buffer unit configured to initialize the reference picture list based on the random access flag, the initializing including ordering reference pictures for the random access picture in a decoded picture buffer, the ordering including assigning an inter-view reference index to each reference picture, the inter-view reference index identifying a reference picture in the reference picture list used for inter-view prediction of a block to be decoded in the random access picture, the assigning the inter-view reference index being performed using the view number information and the view identification information, and obtain modification information for the initialized reference picture list from the multi-view video stream, the modification information representing how to change the inter-view reference index assigned to the reference picture in the initialized reference picture list, and configured to modify the initialized reference picture list for inter-view prediction, the modifying including reordering the reference picture by changing the inter-view reference index assigned to the reference picture based on the modification information; and
an inter-prediction unit configured to determine a prediction value of a current macroblock based on the modified reference picture list, and configured to decode the current macroblock using the prediction value,
wherein the initialization information is obtained from an extension area of a sequence header in the multi-view video stream.
8. The apparatus of claim 7 , wherein the view number information indicates a number of reference views of the random access picture, and the view identification information provides a view identifier of each reference view for the random access picture.
9. The apparatus of claim 7 , wherein the multi-view video data includes video data of a base view independent of other views, the base view being a view decoded without using inter-view prediction.
10. The apparatus of claim 7 , wherein the inter-view reference index is changed by performing a subtraction operation or an addition operation between a view information difference and a predicted inter-view reference index,
wherein the view information difference means a difference between a inter-view reference index of a current reference picture in the initialized reference picture list and the predicted inter-view reference index, and
wherein the predicted inter-view reference index indicates a inter-view reference index of the reference picture assigned right before.
11. The apparatus of claim 7 , wherein the decoded picture buffer unit is configured to shift other remaining pictures in the initialized reference picture list to later positions in the initialized reference picture list.