Moving picture encoding method, moving picture decoding method, moving picture encoding apparatus, and moving picture decoding apparatus
A moving picture encoding method for increasing coding efficiency includes: determining whether or not to apply orthogonal transformation, to calculate a value of an orthogonal transform skip flag; performing the orthogonal transformation on a prediction residual according to the value of the orthogonal transform skip flag, to calculate at least one orthogonal transform coefficient; performing quantization on at least the one orthogonal transform coefficient, to calculate at least one quantized coefficient; performing variable-length encoding on the orthogonal transform skip flag; and changing a scan order for at least the one quantized coefficient according to the value of the orthogonal transform skip flag, and performing variable-length encoding on at least the one quantized coefficient in the scan order after the change.
1. A decoder, comprising:
processing circuitry; and
storage connected to the processing circuitry,
wherein the processing circuitry, in operation of executing programming stored in the storage,
obtains a transform skip flag,
performs variable-length decoding on quantized coefficients of a current block according to a value of the transform skip flag by (i) when the value of the transform skip flag is a first value, (a) determining a first order of applying the variable-length decoding to the quantized coefficients and (b) performing the variable-length decoding according to the first order on the quantized coefficients, and (ii) when the value of the transform skip flag is a second value different from the first value, (a) determining a second order of applying the variable-length decoding to the quantized coefficients, the second order being different from the first order, and (b) performing the variable-length decoding according to the second order on the quantized coefficients,
after performing the variable-length decoding on the quantized coefficients, performs inverse quantization on the quantized coefficients according to the value of the transform skip flag to generate transform coefficients by (i) when the value of the transform skip flag is the first value, performing the inverse quantization on the quantized coefficients using a quantization matrix to generate the transform coefficients and (ii) when the value of the transform skip flag is the second value, performing the inverse quantization on the quantized coefficients without using the quantization matrix to generate the transform coefficients,
obtains a decoded prediction residual of the current block by (i) when the value of the transform skip flag is the first value, performing inverse transform on the transform coefficients to generate the decoded prediction residual of the current block and (ii) when the value of the transform skip flag is the second value, setting the transform coefficients as the decoded prediction residual of the current block, and
generates a reconstructed image using a prediction image and the decoded prediction residual of the current block,
wherein in the determining the first order of applying the variable-length decoding to the quantized coefficients, the processor:
determines whether or not a prediction direction in an intra prediction mode is included in a vertical prediction group;
when the prediction direction in the intra prediction mode is included in the vertical prediction group, a horizontal order is determined as the first order of applying the variable-length decoding to the quantized coefficients; and
when the prediction direction in the intra prediction mode is not included in the vertical prediction group, (i) determines whether or not the prediction direction in the intra prediction mode is included in a horizontal prediction group, (ii) when the prediction direction in the intra prediction mode is included in the horizontal prediction group, a vertical order is determined as the first order of applying the variable-length decoding to the quantized coefficients, and (iii) when the prediction direction in the intra prediction mode is not included in the horizontal prediction group, an order different from the vertical order and the horizontal order is determined as the first order of applying the variable-length decoding to the quantized coefficients.