Video decoding apparatus and video coding apparatus
An image decoding apparatus capable of more preferably applying inverse non-separable transform and techniques related thereto are provided. A video decoding apparatus includes a header decoder configured to decode a flag indicating a high accuracy coding mode from a sequence parameter set SPS, a scaling processing unit configured to perform inverse quantization on a transform coefficient for each transform block, and an inverse transform processing unit configured to perform inverse transform. The scaling processing unit switches whether a variable for indicating a range of the transform coefficient depends on a bit depth or does not depend on the bit depth, based on the flag and a size of the transform block.
1. A video decoding apparatus comprising:
a header decoding circuit configured to decode an extended precision flag in a sequence parameter set, wherein the extend precision flag indicates whether a high bit depth coding mode is used; and
an inverse quantization and inverse transform processing circuit configured to, for each transform unit, perform inverse quantization and perform inverse transform on a transform coefficient, wherein
the header decoding circuit is further configured to derive a transform coefficient range, which is set to one of (i) a value obtained by using a bit depth and (ii) an integer value according to a value of the extended precision flag, and
the inverse quantization and inverse transform processing circuit is further configured to:
derive a first shift value by using the transform coefficient range in a case that a value of a transform skip flag, indicating whether a transform is applied, is equal to zero,
derive the first shift value without using the transform coefficient range in a case that the value of the transform skip flag is equal to one,
derive a transform coefficient value by performing a right shift operation by using the first shift value, and
derive a scaled transform coefficient by performing a clipping operation on the transform coefficient value by using a coefficient minimum value and a coefficient maximum value, wherein the coefficient minimum value and the coefficient maximum value are derived based on the transform coefficient range.
2. A video coding apparatus comprising:
a header coding circuit configured to code an extended precision flag in a sequence parameter set, wherein the extended precision flag indicates whether a high bit depth coding mode is used; and
an inverse quantization and inverse transform processing circuit configured to, for each transform unit, perform inverse quantization and perform inverse transform on a transform coefficient, wherein
the inverse quantization and inverse transform processing circuit is further configured to:
derive a first shift value by using the transform coefficient range in a case that a value of a transform skip flag, indicating whether a transform is applied, is equal to zero,
derive the first shift value without using the transform coefficient range in a case that the value of the transform skip flag is equal to one,
derive a transform coefficient value by performing a right shift operation by using the first shift value, and
derive a scaled transform coefficient by performing a clipping operation on the transform coefficient value by using a coefficient minimum value and a coefficient maximum value, wherein the coefficient minimum value and the coefficient maximum value are derived based on the transform coefficient range, and
the header coding circuit is further configured to derive a transform coefficient range, which is set to one of (i) a value obtained by using a bit depth and (ii) an integer value according to a value of the extended precision flag.