Per-sample prediction encoding apparatus and method

1. A video decoding apparatus comprising:

a splitter configured to split an image into at least one block;

a predictor configured to predict a current sample by using at least one of a value obtained by applying a first weight to a first sample predicted earlier than the current sample in a current block and being adjacent to the current sample in a horizontal direction and a value obtained by applying a first weight to a second sample predicted earlier than the current sample in the current block and being adjacent to the current sample in a vertical direction; and

a decoder configured to decode the image by using a residual value of the current sample obtained from a bitstream and a prediction value of the current sample.

2. The video decoding apparatus of claim 1 , wherein the first weight is proportional to a difference value between the first sample adjacent to the current sample in the horizontal direction and a third sample predicted earlier than the current sample in the current block and being adjacent to the current sample in a diagonal direction.

3. The video decoding apparatus of claim 1 , wherein the second weight is proportional to a difference value between the second sample adjacent to the current sample in the vertical direction and a third sample predicted earlier than the current sample in the current block and being adjacent to the current sample in a diagonal direction.

4. The video decoding apparatus of claim 1 , wherein the first weight and the second weight are equal to each other.

5. The video decoding apparatus of claim 1 , wherein the first sample is located at a boundary of the current sample,

the first sample is predicted by using at least one of a value obtained by applying a fourth weight to a first reference sample outside the current block and adjacent to the first sample in the horizontal direction and a value obtained by applying a fifth weight to a third sample predicted earlier than the current sample in the current block and being adjacent to the current sample in a diagonal direction,

the fourth weight is proportional to a difference value between the first reference sample and a second reference sample outside the current block and adjacent to the first sample in a diagonal direction, and

the fifth weight is proportional to a difference value between the third sample and the second reference sample.

6. The video decoding apparatus of claim 1 , wherein the second sample is located at a boundary of the current sample,

the second sample is predicted by using at least one of a value obtained by applying a fourth weight to a third sample predicted earlier than the current sample in the current block and being adjacent to the current sample in a diagonal direction and a value obtained by applying a fifth weight to a first reference sample outside the current block and adjacent to the second sample in a vertical direction,

the fourth weight is proportional to a difference value between the third sample and a second reference sample outside the current block and adjacent to the second sample in a diagonal direction, and

the fifth weight is proportional to a difference value between the first reference sample and the second reference sample.

7. The video decoding apparatus of claim 1 , wherein a third sample predicted earlier than the current sample in the current block and being adjacent to the current sample in a diagonal direction is located at a boundary of the current block,

the third sample is obtained by using at least one of a value obtained by applying a fourth weight to a first reference sample outside the current block and adjacent to the third sample in a horizontal direction and a value obtained by applying a fifth weight to a second reference sample outside the current block and adjacent to the third sample in a vertical direction,

the fourth weight is proportional to a difference value between the first reference sample and a third reference sample adjacent to the third sample in a diagonal direction, and

the fifth weight is proportional to a difference value between the second reference sample and the third reference sample.

8. A video decoding method comprising:

splitting an image into at least one block;

predicting a current sample by using at least one of a value obtained by applying a first weight to a first sample predicted earlier than the current sample in a current block and being adjacent to the current sample in a horizontal direction and a value obtained by applying a second weight to a second sample predicted earlier than the current sample in the current block and being adjacent to the current sample in a vertical direction; and

decoding the image by using a residual value of the current sample obtained from a bitstream and a prediction value of the current sample.