Method and apparatus for encoding/decoding multilayer video signal

A method for decoding a multi-layer video signal with a decoding apparatus, the method comprising:

• obtaining, a maximum temporal level identifier relating to a lower layer, the maximum temporal level identifier indicating a maximum temporal level identifier to which an inter-layer prediction of an upper layer is allowed;
• determining, whether a corresponding picture of the lower layer is used as an inter-layer reference picture for a current picture of the upper layer by comparing the obtained maximum temporal level identifier with a temporal level identifier of the corresponding picture of the lower layer;
• and
• performing, the inter-layer prediction of the current picture using the corresponding picture of the lower layer,
• wherein a tile alignment present flag is obtained from a bitstream when video usability information is presented in the bitstream, the tile alignment present flag equal to 1 indicates that a first restriction is applied, and the tile alignment present flag equal to 0 indicates that the first restriction may or may not be applied,
• wherein the first restriction is to restrict all pictures belonging to a video sequence to consist of a single tile,
• wherein a tile boundary alignment flag is obtained, based on the tile alignment present flag, from the bitstream, and the tile boundary alignment flag indicates whether a second restriction is applied,
• wherein the second restriction is to restrict i) two samples of a corresponding picture of the lower layer to belong to one tile when two samples of the current picture of the upper layer belong to one tile, and ii) the two samples of the corresponding picture of the lower layer to belong to different tiles when the two samples of the current picture of the upper layer belong to different tiles,
• wherein a 8-tap filter is applied to the corresponding picture belonging to the lower layer for performing inter-layer prediction of luminance components of the current picture,
• wherein coefficient sets of the 8-tap filter include {−1, 4, −10, 31, 47, −9, 3, −1}, and
• wherein when the corresponding picture is used for inter-layer prediction of a current block of the current picture, both x and y components of a motion vector relating to the current block are derived equal to zero.

A method for encoding a multi-layer video signal with an encoding apparatus, the method comprising:

• encoding a tile alignment present flag indicating whether a first restriction is applied, wherein the first restriction is to restrict all pictures belonging to a video sequence to consist of a single tile;
• encoding, based on the tile alignment present flag, a tile boundary alignment flag indicating whether a second restriction is applied, wherein the second restriction is to restrict i) two samples of a corresponding picture of the lower layer to belong to one tile when two samples of the current picture of the upper layer belong to one tile, and ii) the two samples of the corresponding picture of the lower layer to belong to different tiles when the two samples of the current picture of the upper layer belong to different tiles;
• and
• generating a bitstream including at least one of the tile alignment present flag and the tile boundary alignment flag,
• wherein the corresponding picture of the lower layer is a picture to which a re-sampling filter is applied for performing inter-layer prediction on the current picture, and
• wherein when the corresponding picture is used for inter-layer prediction of a current block of the current picture, both x and y components of a motion vector relating to the current block are derived equal to zero.