Method and device for processing a video signal using inter-view prediction
A method and apparatus for processing a video signal using inter-viewpoint prediction are disclosed. The video signal processing method includes acquiring a type of a current block, acquiring a depth inter-view prediction flag according to the current block type, acquiring color information of a reference block corresponding to the current block when the current block is encoded using depth inter-view prediction according to the depth inter-view prediction flag, and predicting color information of the current block using the color information of the reference block. The video signal processing method performs inter-viewpoint prediction using depth information, such that a motion vector of a macroblock can be induced from a decoder using such depth information. Therefore, a motion vector of the macroblock is not separately coded. In addition, the video signal processing method uses a flag that indicates whether to perform inter-viewpoint prediction using depth information.
1. A method for processing a video signal, the method comprising:
receiving a video signal, the video signal including a current color picture and a current depth picture;
acquiring a depth inter-view prediction flag for a current block in the current color picture, the depth inter-view prediction flag indicating whether the current block is coded using depth inter-view prediction, the depth inter-view prediction being used to predict color information of the current block from a reference block in a neighboring view of the current block by using depth information of a corresponding depth block in the current depth picture;
obtaining motion skip flag information corresponding the current color block from the video signal;
acquiring color information of the reference block corresponding to the current block based on the motion skip flag information when the current block is coded using the depth inter-view prediction according to the depth inter-view prediction flag; and
predicting the color information of the current block using the color information of the reference block,
wherein, based on the motion skip flag information indicating that the current color block is using motion skip, acquiring the color information of the reference block comprises:
obtaining a depth skip indication flag from the video signal, the depth skip indication flag indicating whether the current block is coded in a depth skip mode;
acquiring the depth information of the corresponding depth block for the current block based on the depth skip indication flag;
deriving disparity vectors for pixels of the current block based on a value resulting from dividing a product of a focal length of a current viewpoint camera and a distance between the current viewpoint camera and a reference viewpoint camera by the depth information of the corresponding depth block;
deriving a representative disparity vector for the current block from an average value of the derived disparity vectors; and
determining the reference block corresponding to the current block based on the derived representative disparity vector,
wherein the depth information indicates a disparity difference between the current viewpoint camera and the reference viewpoint camera,
wherein, based on the motion skip flag information indicating that the current color block is not using motion skip, acquiring the color information of the reference block comprises:
obtaining a global motion vector based on inter-view reference information;
obtaining additional information including offset information and viewpoint direction information;
obtaining a corrected global motion vector based on the global motion vector and the offset information; and
acquiring the color information of the reference block based on the corrected global motion vector and the viewpoint direction information.
2. The method according to claim 1 , wherein the disparity vector is derived in units of a pixel of the corresponding depth block.
3. The method according to claim 1 , wherein the corresponding depth block is a block of the current depth picture located at a same position as the current block.
4. The method according to claim 1 , wherein the disparity vector represents a disparity between a pixel position of the current block and a pixel position of the reference block.