Intra-prediction mode-based image processing method and apparatus therefor

1. A method of processing an image based on an intra prediction mode, the method comprising:

generating a first prediction sample by performing a spatial prediction on a current chroma block;

determining whether a combined intra prediction mode is applied to the current chroma block, wherein the combined intra prediction mode indicates a prediction mode in which a spatial prediction method and an inter component prediction method have been combined;

determining a component prediction mode used for an inter component prediction of the current chroma block when the combined intra prediction mode is applied to the current chroma block, wherein the component prediction mode indicates a prediction mode for generating a prediction sample of a chroma component using a sample value of a luma component;

generating a second prediction sample by applying the determined component prediction mode to the current chroma block; and

generating a final prediction sample of the current chroma block using the first prediction sample and the second prediction sample.

2. The method of claim 1 ,

wherein determining the component prediction mode comprises determining the component prediction mode among a first prediction mode using one linear model or a second prediction mode using a plurality of linear models.

3. The method of claim 2 ,

wherein generating the second prediction sample comprises generating the second prediction sample using a reconstructed sample of a luma block corresponding to the current chroma block when the determined component prediction mode is the first prediction mode.

4. The method of claim 2 ,

wherein generating the second prediction sample comprises classifying reconstructed samples of a luma block, corresponding to the current chroma block, into two groups when the determined component prediction mode is the second prediction mode, and

wherein the second prediction sample is generated using a reconstructed sample of a luma block corresponding to the current chroma block in the classified group unit.

5. The method of claim 4 ,

wherein classifying the reconstructed samples into the two groups comprises deriving a threshold using neighbor samples of a luma block corresponding to the current chroma block, and

wherein the reconstructed samples of the luma block are classified into the two groups based on the derived threshold.

6. The method of claim 1 ,

wherein determining the component prediction mode comprises extracting an index indicating a specific mode among the first prediction mode or the second prediction mode.

7. The method of claim 6 ,

wherein extracting the index indicating the specific mode comprises identifying whether the current chroma block is greater than a predetermined size, and

wherein the index indicating the specific mode is extracted when the current chroma block is greater than the predetermined size as a result of the identification.

8. The method of claim 1 ,

wherein determining whether the combined intra prediction mode is applied comprises extracting a combined intra prediction flag indicating whether the combined intra prediction mode is applied to the current chroma block, and

wherein a plurality of context models is assigned to the combined intra prediction flag.

9. The method of claim 8 ,

wherein a context model of the combined intra prediction flag is determined based on whether the combined intra prediction mode is applied to a block neighboring the current chroma block.

10. The method of claim 1 ,

wherein generating the final prediction sample comprises generating the final prediction sample by applying predetermined weights to the first prediction sample and the second prediction sample.

11. The method of claim 1 ,

wherein generating the final prediction sample comprises determining a weight applied to each of the first prediction sample and the second prediction sample, and

wherein the final prediction sample is generated by applying the determined weights to the first prediction sample and the second prediction sample.

12. The method of claim 11 ,

wherein generating the final prediction sample comprises extracting an index indicating a specific weight among a plurality of predetermined weights.

13. The method of claim 11 ,

wherein the weights applied to the first prediction sample and the second prediction sample are determined based on a prediction direction of a prediction mode used for the spatial prediction and a distance between a reference sample neighboring the current chroma block and a sample within the current chroma block.

14. An apparatus for processing an image based on an intra prediction mode, the apparatus comprising:

a first prediction sample generation unit configured to generate a first prediction sample by performing a spatial prediction on a current chroma block;

a combined intra prediction mode determination unit configured to determine whether a combined intra prediction mode is applied to the current chroma block, wherein the combined intra prediction mode indicates a prediction mode in which a spatial prediction method and an inter component prediction method have been combined;

a component prediction mode determination unit configured to determine a component prediction mode used for an inter component prediction of the current chroma block when the combined intra prediction mode is applied to the current chroma block, wherein the component prediction mode indicates a prediction mode for generating a prediction sample of a chroma component using a sample value of a luma component;

a second prediction sample generation unit configured to generate a second prediction sample by applying the determined component prediction mode to the current chroma block; and

a final prediction sample generation unit configured to generate a final prediction sample of the current chroma block using the first prediction sample and the second prediction sample.