Image decoding device, image decoding method, and program

An image decoding method, comprising:

• (A) decoding and outputting prediction information and quantized values;
• (B) subjecting the quantized values outputted in the (A) to inverse quantization processing to generate transformation coefficients and outputting the transformation coefficients;
• (C) subjecting the transformation coefficients outputted in the (B) to inverse transformation processing to generate prediction residuals and outputting the prediction residuals;
• (D) generating predicted pixels on the basis of the prediction information outputted in the (A) and outputting the predicted pixels;
• (E) adding the prediction residuals outputted in the step (C) and the predicted pixels outputted in the (D) to acquire decoded pixels and outputting the decoded pixels;
• (F) accumulating the decoded pixels outputted in the (E);
• (G) generating predicted pixels on the basis of the decoded pixels outputted in the (E) and the prediction information outputted in the (A) and outputting the predicted pixels;
• and
• (H) generating predicted pixels on the basis of the decoded pixels accumulated in the (F) and the prediction information outputted in the (A) and outputting the predicted pixels, wherein
• in the (D), the predicted pixels are generated by performing weighted averaging processing using weighting factors on the predicted pixels outputted in the (G) and the predicted pixels outputted in the (H),
• in the (D), in each of a plurality of small regions obtained by dividing a unit block of a decoding target block, the predicted pixels generated in the (G) or/and the predicted pixels generated in the (H) is employed, and
• in the (D), when intra prediction is allocated to the small region, a prediction mode according to a division configuration of the small regions is used.

A program stored on a non-transitory computer-readable medium for causing a computer to function as an image decoding device,

• the image decoding device including a circuit, wherein
• the circuit:
• decodes prediction information and quantized values;
• subjects the quantized values to inverse quantization processing to generate transformation coefficients;
• subjects the transformation coefficients to inverse transformation processing to generate prediction residuals;
• generates predicted pixels on the basis of the prediction information;
• adds the prediction residuals and the predicted pixels to acquire decoded pixels;
• accumulates the decoded pixels;
• generates first predicted pixels on the basis of the decoded pixels and the prediction information;
• and
• generates second predicted pixels on the basis of the accumulated decoded pixels a and the prediction information, wherein
• the circuit generates the predicted pixels by performing weighted averaging processing using weighting factors on the first predicted pixels and the second predicted pixels,
• the circuit employs, in each of a plurality of small regions obtained by dividing a unit block of a decoding target block, the first predicted pixels or/and the second predicted pixels, and
• the circuit, when intra prediction is allocated to the small region, uses a prediction mode according to a division configuration of the small regions.

An image decoding device comprising a circuit, wherein

• the circuit:
• decodes prediction information and quantized values;
• subjects the quantized values to inverse quantization processing to generate transformation coefficients;
• subjects the transformation coefficients to inverse transformation processing to generate prediction residuals;
• generates predicted pixels on the basis of the prediction information;
• adds the prediction residuals and the predicted pixels to acquire decoded pixels;
• accumulates the decoded pixels;
• generates first predicted pixels on the basis of the decoded pixels and the prediction information;
• and
• generates second predicted pixels on the basis of the accumulated decoded pixels and the prediction information, wherein
• the circuit generates the predicted pixels by performing weighted averaging processing using weighting factors on the first predicted pixels and the second predicted pixels,
• the circuit employs, in each of a plurality of small regions obtained by dividing a unit block of a decoding target block, the first predicted pixels or/and the second predicted pixels, and
• the circuit, when intra prediction is allocated to the small region, uses a prediction mode according to a type of division of the small regions.