# US 2020/0404264 A1 · Method and image processing apparatus for video coding

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2020/0404264 A1
- Document ID: US20200404264A1
- Status: Published application
- Kind: A1
- Application number: 16/910,068
- Filed: 2020-06-24
- Published: 2020-12-24
- Earliest claimed: 2019-06-24
- Assignees: Industrial Technology Research Institute
- Applicants: Industrial Technology Research Institute
- Inventors: Chang-Hao Yau; Bing-Jung Fuh; Ching-Chieh Lin
- Web page: https://patentgrep.com/patent/US20200404264A1

## CPC

- H04N 19/117: Filters, e.g. for pre-processing or post-processing (sub-band filter banks H04N19/635)
- H04N 19/134: characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N 19/157: Assigned coding mode, i.e. the coding mode being predefined or preselected to be further used for selection of another element or parameter
- H04N 19/176: the region being a block, e.g. a macroblock
- H04N 19/59: involving spatial sub-sampling or interpolation, e.g. alteration of picture size or resolution
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

A method and an image processing apparatus for video coding are provided. The method is adapted to the image processing apparatus and includes following steps. A current coding unit is received, and filter selection is performed according to a size of the current coding unit. At least one selected filter is used to perform a filtering operation on a plurality of reference boundaries of the current coding unit to generate a plurality of filtering reference values. An interpolation operation is performed on the current coding unit according to the filtering reference values to generate a plurality of interpolated prediction values.

## Claims

### Claim 1

A method for video coding, adapted for an image processing apparatus, and comprising:
receiving a current coding unit;
performing filter selection according to a size of the current coding unit;
using at least one selected filter to perform a filtering operation on a plurality of reference boundaries of the current coding unit to generate a plurality of filtering reference values; and
performing an interpolation operation on the current coding unit according to the filtering reference values to generate a plurality of interpolated prediction values.

### Claim 2 (depends on 1)

The method for video coding as claimed in claim 1, wherein the step of performing filter selection according to the size of the current coding unit comprises:
performing parameter selection according to the size of the current coding unit and a boundary size of each of the reference boundaries to generate at least one filter index of the current coding unit, wherein each of the at least one filter index corresponds to one of the plurality of filters.

### Claim 3 (depends on 2)

The method for video coding as claimed in claim 2, wherein the step of performing parameter selection according to the size of the current coding unit and the boundary size of each of the reference boundaries to generate the at least one filter index of the current coding unit comprises:
calculating a combination of each of the reference boundaries of the current coding unit according to the size of the current coding unit;
calculating a split size of the current coding unit according to the boundary size and the combination of each of the reference boundaries of the current coding unit; and
calculating the at least one filter index of the current coding unit according to the split size of the current coding unit.

### Claim 4 (depends on 3)

The method for video coding as claimed in claim 3, wherein the step of calculating the combination of each of the reference boundaries of the current coding unit according to the size of the current coding unit comprises:
determining whether a width and a height of the current coding unit are both 4;
if yes, setting the combination of each of the reference boundaries of the current coding unit to 2; and
if not, setting the combination of each of the reference boundaries of the current coding unit to 4.

### Claim 5 (depends on 3)

The method for video coding as claimed in claim 3, wherein the step of calculating the split size of the current coding unit according to the boundary size and the combination of each of the reference boundaries of the current coding unit comprises:
performing a dividing operation on the boundary size and the combination of each of the reference boundaries of the current coding unit to generate the split size of the current coding unit.

### Claim 6 (depends on 3)

The method for video coding as claimed in claim 3, wherein the step of calculating the at least one filter index of the current coding unit according to the split size of the current coding unit comprises:
performing a logarithm operation on the split size of the current coding unit to generate the at least one filter index of the current coding unit.

### Claim 7 (depends on 1)

The method for video coding as claimed in claim 1, wherein the selected filter is selected from a plurality of smooth filters of different taps.

### Claim 8 (depends on 7)

The method for video coding as claimed in claim 7, wherein smooth filters of different taps comprise a bypass filter and 2N-tap mean filter, wherein N is 1, 2, 3 and 4.

### Claim 9 (depends on 1)

The method for video coding as claimed in claim 1, wherein the reference boundaries are a top reference row and a left reference column of the current coding unit, and the step of using the at least one selected filter to perform the filtering operation on each of the reference boundaries of the current coding unit to generate the filtering reference values comprises:
using each of the at least one selected filter to respectively perform the filtering operation on the top reference row and the left reference column of the current coding unit to respectively generate the filtering reference values.

### Claim 10 (depends on 1)

The method for video coding as claimed in claim 1, wherein the step of performing the interpolation operation on the current coding unit according to the filtering reference values to generate the interpolated prediction values comprises:
calculating a plurality of down-sampling prediction values of the current coding unit according to the filtering reference values, a preset matrix, and a preset offset vector to generate a down-sampling prediction sample; and
generating the interpolated prediction values according to a plurality of original edge values of the reference boundaries and the down-sampling prediction sample.

### Claim 11 (depends on 10)

The method for video coding as claimed in claim 10, wherein the preset matrix and the preset offset vector are related to the size of the current coding unit.

### Claim 12 (depends on 10)

The method for video coding as claimed in claim 10, wherein the step of generating the interpolated prediction values according to the original edge values of the reference boundaries and the down-sampling prediction sample comprises:
performing up-sampling processing on the down-sampling prediction sample to generate a prediction sample of the current coding unit, wherein each of the down-sampling prediction values is equidistantly arranged in the prediction sample;
arranging the original edge values of the reference boundaries to a position corresponding to a top row and a left column of the prediction sample;
performing the interpolation operation on the prediction sample according to the original edge values and the down-sampling prediction values to generate an interpolated sample prediction value of the prediction sample; and
setting the down-sampling prediction values and the interpolated sample prediction value in the prediction sample as the interpolated prediction values.

### Claim 13 (depends on 12)

The method for video coding as claimed in claim 12, wherein the step of performing the interpolation operation on the prediction sample according to the original edge values and the down-sampling prediction values of the reference boundaries comprises:
dividing the prediction sample into a plurality of interpolation modules according to the size of the current coding unit and a size of the down-sampling prediction sample; and
performing linear interpolation on each of the interpolation modules by using the original edge values and the down-sampling prediction values to generate the interpolated sample prediction value of the prediction sample.

### Claim 14 (depends on 13)

The method for video coding as claimed in claim 13, wherein regarding each of the interpolation modules, a most lower right small block in internal of the each of the interpolation modules, and a lower left small block, an upper left small block and an upper right small block located outside the each of the interpolation modules are four known small blocks, and are composed of four of the original edge values and the down-sampling prediction values.

### Claim 15 (depends on 12)

The method for video coding as claimed in claim 12, wherein the step of performing the interpolation operation on the prediction sample according to the original edge values and the down-sampling prediction values of the reference boundaries comprises:
dividing the prediction sample into a plurality of horizontal interpolation modules according to the size of the current coding unit and a size of the down-sampling prediction sample, wherein a height of each of the horizontal interpolation modules is 1;
performing linear interpolation on each of the horizontal interpolation modules by using the original edge values and the down-sampling prediction values to generate a plurality of horizontal interpolated sample prediction values of the prediction sample;
dividing the prediction sample into a plurality of vertical interpolation modules according to the size of the current coding unit and the size of the down-sampling prediction sample, wherein a width of each of the vertical interpolation modules is 1;
performing linear interpolation on each of the vertical interpolation modules by using the original edge values, the down-sampling prediction values and the horizontal interpolated sample prediction values to generate a plurality of vertical interpolated sample prediction values of the prediction sample; and
setting the down-sampling prediction values, the horizontal interpolated sample prediction values and the vertical interpolated sample prediction values in the prediction sample as the interpolated prediction values.

### Claim 16 (depends on 15)

The method for video coding as claimed in claim 15, wherein
regarding each of the horizontal interpolation modules, a rightmost small block in internal of the each of the horizontal interpolation modules, and a left small block located outside the each of the horizontal interpolation modules are composed of two of the original edge values and the down-sampling prediction values; and
regarding each of the vertical interpolation modules, a lowermost small block in internal of the each of the vertical interpolation modules, and an upper small block located outside the each of the vertical interpolation modules are composed of two of the original edge values, the down-sampling prediction values and the horizontal interpolated sample prediction values.

### Claim 17 (depends on 12)

The method for video coding as claimed in claim 12, wherein the step of performing the interpolation operation on the prediction sample according to the original edge values and the down-sampling prediction values of the reference boundaries comprises:
dividing the prediction sample into a plurality of vertical interpolation modules according to the size of the current coding unit and a size of the down-sampling prediction sample, wherein a width of each of the vertical interpolation modules is 1;
performing linear interpolation on each of the vertical interpolation modules by using the original edge values and the down-sampling prediction values to generate a plurality of vertical interpolated sample prediction values of the prediction sample;
dividing the prediction sample into a plurality of horizontal interpolation modules according to the size of the current coding unit and the size of the down-sampling prediction sample, wherein a height of each of the horizontal interpolation modules is 1;
performing linear interpolation on each of the horizontal interpolation modules by using the original edge values, the down-sampling prediction values and the vertical interpolated sample prediction values to generate a plurality of horizontal interpolated sample prediction values of the prediction sample; and
setting the down-sampling prediction values, the horizontal interpolated sample prediction values and the vertical interpolated sample prediction values in the prediction sample as the interpolated prediction values.

### Claim 18

An image processing apparatus, comprising:
a memory, configured to store data; and
a processor, coupled to the memory, and configured to:
receive a current coding unit;
perform filter selection according to a size of the current coding unit;
use at least one selected filter to perform a filtering operation on a plurality of reference boundaries of the current coding unit to generate a plurality of filtering reference values; and
perform an interpolation operation on the current coding unit according to the filtering reference values to generate a plurality of interpolated prediction values.

### Claim 19 (depends on 18)

The image processing apparatus as claimed in claim 18, wherein the processor performs parameter selection according to the size of the current coding unit and a boundary size of each of the reference boundaries to generate at least one filter index of the current coding unit, wherein each of the at least one filter index corresponds to one of a plurality of filters.

### Claim 20 (depends on 18)

The image processing apparatus as claimed in claim 18, wherein the processor calculates a plurality of down-sampling prediction values of the current coding unit according to the filtering reference values, a preset matrix, and a preset offset vector to generate a down-sampling prediction sample, and generates the interpolated prediction values according to a plurality of original edge values of the reference boundaries and the down-sampling prediction sample.
