Method for processing image on basis of transform and device therefor
Embodiments of the present disclosure provide a method and apparatus for processing a video signal. A method for decoding a video signal according to an embodiment of the present disclosure includes obtaining a sub-block transform (SBT) flag indicating whether an SBT is applied, wherein the SBT represents a transform applied to one of sub-blocks split from a coding unit, determining a transform unit from the coding unit based on the SBT flag, determining a size value of the transform unit as a first reference value if DST-7 (Discrete Sine Transform type 7) or DCT-8 (Discrete Cosine Transform type 8) is applied and the size value of the transform unit is greater than the first reference value and less than a third reference value and determining the size value as a second reference value if DCT-2 (Discrete Cosine Transform type 2) is applied and the size value is equal to or greater than the third reference value, wherein the third reference value is greater than the second reference value and the second reference value is greater than the first reference value, obtaining transform coefficients based on the size value, and performing an inverse transform on the transform coefficients. A data processing time and the amount of data necessary for a transform can be reduced by performing coding in consideration of a region reduced according to a block size.
1. A method for decoding a video signal, comprising:
obtaining a sub-block transform (SBT) flag indicating whether a SBT is applied, wherein the SBT represents a transform applied to a transform unit corresponding to one of subblocks split from a coding unit;
checking whether a condition related to a width and a height of the transform unit is satisfied;
determining a width variable and a height variable related to a transform target region based on the SBT flag and the condition; and
performing an inverse-transform on the transform target region based on the width variable and the height variable related to the transform target region,
wherein, based on the SBT flag indicating that the SBT is applied to the transform unit and the condition being satisfied, the width variable related to the transform target region is determined as a first reference value,
wherein, based on the SBT flag indicating that the SBT is not applied to the transform unit or the condition not being satisfied, the width variable related to the transform target region is determined as a smaller value among the width of the transform unit and a second reference value,
wherein, based on the SBT flag indicating that the SBT is applied to the transform unit and the condition being satisfied, the height variable related to the transform target region is determined as the first reference value,
wherein, based on the SBT flag indicating that the SBT is not applied to the transform unit or the condition not being satisfied, the height variable related to the transform target region is determined as a smaller value among the height of the transform unit and the second reference value,
wherein the inverse-transform on the transform target region comprises a horizontal direction transform and a vertical direction transform on a non-zero region in the transform unit,
wherein a horizontal transform kernel index indicating a type of a horizontal transform kernel for the horizontal direction transform is derived based on the SBT flag,
wherein a vertical transform kernel index indicating a type of a vertical transform kernel for the vertical direction transform is derived based on the SBT flag,
wherein a width of the non-zero region is determined as a smaller value among the width of the transform unit and 16, based on the horizontal transform kernel index being greater than zero, and
wherein a height of the non-zero region is determined as a smaller value among the height of the transform unit and 16, based on the vertical transform kernel index being greater than zero.
2. A method for encoding a video signal, the method comprising:
checking whether a condition related to a width and a height of a transform unit is satisfied;
determining a width variable and a height variable related to a transform target region based on whether a sub-block transform (SBT) is applied and the condition;
performing a transform on the transform unit based on the width variable and the height variable related to the transform target region; and
generating a SBT flag indicating whether the SBT is applied, wherein the SBT represents a transform applied to the transform unit corresponding to one of subblocks split from a coding unit,
wherein, based on the SBT being applied to the transform unit and the condition being satisfied, the width variable related to the transform target region is determined as a first reference value,
wherein, based on the SBT not being applied to the transform unit or the condition not being satisfied, the width variable related to the transform target region is determined as a smaller value among the width of the transform unit and a second reference value, and
wherein, based on the SBT being applied to the transform unit and the condition being satisfied, the height variable related to the transform target region is determined as the first reference value,
wherein, based on the SBT not being applied to the transform unit or the condition not being satisfied, the height variable related to the transform target region is determined as a smaller value among the height of the transform unit and the second reference value,
wherein the transform on the transform unit comprises a horizontal direction transform and a vertical direction transform on the transform unit,
wherein a horizontal transform kernel index indicating a type of a horizontal transform kernel for the horizontal direction transform is derived based on the SBT flag,
wherein a vertical transform kernel index indicating a type of a vertical transform kernel for the vertical direction transform is derived based on the SBT flag,
wherein a width of an area, in which a non-zero transform coefficient is present, is determined as a smaller value among the width of the transform unit and 16, based on the horizontal transform kernel index being greater than zero, and
wherein a height of the area, in which the non-zero transform coefficient is present, is determined as a smaller value among the height of the transform unit and 16, based on the vertical transform kernel index being greater than zero.
3. A method comprising:
generating a bitstream for image information including a sub-block transform (SBT) flag indicating whether a SBT is applied, wherein the SBT represents a transform applied to a transform unit corresponding to one of subblocks split from a coding unit; and
transmitting data including the bitstream,
wherein whether a condition related to a width and a height of the transform unit is satisfied is checked,
wherein a width variable and a height variable related to a transform target region is determined based on whether the SBT is applied and the condition,
wherein a transform on the transform unit is performed based on the width variable and the height variable related to the transform target region,
wherein, based on the SBT being applied to the transform unit and the condition being satisfied, the width variable related to the transform target region is determined as a first reference value,
wherein, based on the SBT not being applied to the transform unit or the condition not being satisfied, the width variable related to the transform target region is determined as a smaller value among the width of the transform unit and a second reference value, and
wherein, based on the SBT being applied to the transform unit and the condition being satisfied, the height variable related to the transform target region is determined as the first reference value,
wherein, based on the SBT not being applied to the transform unit or the condition not being satisfied, the height variable related to the transform target region is determined as a smaller value among the height of the transform unit and the second reference value,
wherein the transform on the transform unit comprises a horizontal direction transform and a vertical direction transform on the transform unit,
wherein a horizontal transform kernel index indicating a type of a horizontal transform kernel for the horizontal direction transform is derived based on the SBT flag,
wherein a vertical transform kernel index indicating a type of a vertical transform kernel for the vertical direction transform is derived based on the SBT flag,
wherein a width of an area, in which a non-zero transform coefficient is present, is determined as a smaller value among the width of the transform unit and 16, based on the horizontal transform kernel index being greater than zero, and
wherein a height of the area, in which the non-zero transform coefficient is present, is determined as a smaller value among the height of the transform unit and 16, based on the vertical transform kernel index being greater than zero.