# US 2026/0303859 A1 · Inter prediction based image coding method and apparatus therefor

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

- Number: US 2026/0303859 A1
- Document ID: US20260303859A1
- Status: Published application
- Kind: A1
- Application number: 19/578,565
- Filed: 2026-03-25
- Published: 2026-10-01
- Earliest claimed: 2025-03-26
- Assignees: Lx Semicon Co., Ltd.
- Applicants: Lx Semicon Co., Ltd.
- Inventors: Jin Wook LEE; Young Jun Kim
- Web page: https://patentgrep.com/patent/US20260303859A1

## CPC

- H04N 19/577: Motion compensation with bidirectional frame interpolation, i.e. using B-pictures
- H04N 19/105: Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction
- H04N 19/132: Sampling, masking or truncation of coding units, e.g. adaptive resampling, frame skipping, frame interpolation or high-frequency transform coefficient masking
- H04N 19/139: Analysis of motion vectors, e.g. their magnitude, direction, variance or reliability
- H04N 19/172: the region being a picture, frame or field
- H04N 19/176: the region being a block, e.g. a macroblock
- H04N 19/52: by predictive encoding
- H04N 19/80: Details of filtering operations specially adapted for video compression, e.g. for pixel interpolation (H04N19/635, H04N19/86 take precedence)
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

An image decoding method according to an embodiment of the disclosure includes obtaining prediction related information for a current block in a current frame, generating a Temporally Interpolated Prediction (TIP) reference frame by interpolating reference frames of the current frame, deriving motion information of the current block based on the prediction related information, and generating prediction samples for the current block based on the motion information.

## Claims

### Claim 1

An image decoding method performed by a decoding apparatus, the image decoding method comprising:
obtaining prediction related information for a current block in a current frame;
generating a Temporally Interpolated Prediction (TIP) reference frame by interpolating reference frames of the current frame;
deriving motion information of the current block based on the prediction related information; and
generating prediction samples for the current block based on the motion information.

### Claim 2 (depends on 1)

The image decoding method of claim 1, wherein the reference frames include a backward reference frame and a forward reference frame of the current frame.

### Claim 3 (depends on 2)

The image decoding method of claim 2, wherein a first temporal distance between the backward reference frame and the current frame is the same as a second temporal distance between the forward reference frame and the current frame.

### Claim 4 (depends on 3)

The image decoding method of claim 3, wherein the first temporal distance and the second temporal distance are preset values.

### Claim 5 (depends on 3)

The image decoding method of claim 3, further comprising: obtaining TIP reference frame related information for the current frame,
wherein the TIP reference frame related information includes temporal distance information indicating the first temporal distance and the second temporal distance.

### Claim 6 (depends on 2)

The image decoding method of claim 2, wherein a first temporal distance between the backward reference frame and the current frame is different from a second temporal distance between the forward reference frame and the current frame.

### Claim 7 (depends on 6)

The image decoding method of claim 6, further comprising: obtaining TIP reference frame related information for the current frame,
wherein the TIP reference frame related information includes a reference frame index indicating the backward reference frame and a reference frame index indicating the forward reference frame.

### Claim 8 (depends on 6)

The image decoding method of claim 6, wherein the TIP reference frame is derived by interpolating the backward reference frame and the forward reference frame based on the first temporal distance and the second temporal distance.

### Claim 9 (depends on 2)

The image decoding method of claim 2, wherein a reference sample of the TIP reference frame is derived by interpolating a reference sample of the backward reference frame and a reference sample of the forward reference frame, and
the reference sample of the forward reference frame is a reference sample at a position moved from a position of the reference sample of the backward reference frame in the forward reference frame based on a temporal motion vector.

### Claim 10 (depends on 9)

The image decoding method of claim 9, wherein the temporal motion vector is derived based on a stored motion vector for the backward reference frame.

### Claim 11 (depends on 9)

The image decoding method of claim 9, further comprising: obtaining TIP reference frame related information for the current frame,
wherein the TIP reference frame related information includes information about the temporal motion vector.

### Claim 12 (depends on 1)

The image decoding method of claim 1, further comprising: obtaining TIP reference frame related information for the current frame,
wherein the TIP reference frame related information includes a first reference frame index indicating a first reference frame and a second reference frame index indicating a second reference frame, and
the reference frames include the first reference frame and the second reference frame of the current frame.

### Claim 13 (depends on 12)

The image decoding method of claim 12, wherein the TIP reference frame is derived by interpolating the first reference frame and the second reference frame based on a first temporal distance and a second temporal distance,
the first temporal distance is a temporal distance between the first frame and the current frame, and
the second temporal distance is a temporal distance between the second frame and the current frame.

### Claim 14 (depends on 12)

The image decoding method of claim 12, wherein the TIP reference frame is derived by interpolating the first reference frame and the second reference frame based on a first interpolation filter coefficient and a second interpolation filter coefficient, and
the TIP reference frame related information includes filter coefficient information for the first interpolation filter coefficient and the second interpolation filter coefficient.

### Claim 15 (depends on 1)

The image decoding method of claim 1, wherein the prediction related information includes a reference frame index indicating the TIP reference frame, and
when a value of the reference frame index is a specific value, the reference frame index indicates the TIP reference frame.

### Claim 16 (depends on 15)

The image decoding method of claim 15, wherein, when the value of the reference frame index is the specific value, a TIP mode for performing prediction based on the TIP reference frame is applied to the current block.

### Claim 17 (depends on 15)

The image decoding method of claim 15, wherein the specific value is a preset value.

### Claim 18

An image encoding method performed by an encoding apparatus, the image encoding method comprising:
generating a Temporally Interpolated Prediction (TIP) reference frame by interpolating reference frames of a current frame;
deriving motion information of a current block in the current frame;
generating prediction samples for the current block based on a reference sample in the TIP reference frame derived based on the motion information; and
encoding image information including prediction related information for the current block.

### Claim 19 (depends on 18)

The image encoding method of claim 18, wherein the reference frames include a backward reference frame and a forward reference frame of the current frame.

### Claim 20

A transmission method for image data, comprising:
obtaining a bitstream generated by an image encoding method, wherein the image encoding method comprises generating a Temporally Interpolated Prediction (TIP) reference frame by interpolating reference frames of a current frame, deriving motion information of a current block in the current frame, generating prediction samples for the current block based on a reference sample in the TIP reference frame derived based on the motion information, and encoding image information including prediction related information for the current block; and
transmitting image data including the bitstream.
