# US 2026/0101033 A1 · Encoding and decoding methods using l-shaped partitions and corresponding apparatuses

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

- Number: US 2026/0101033 A1
- Document ID: US20260101033A1
- Status: Published application
- Kind: A1
- Application number: 19/138,632
- Filed: 2023-11-27
- Published: 2026-04-09
- Earliest claimed: 2022-12-13
- Assignees: InterDigital CE Patent Holdings, SAS
- Applicants: InterDigital CE Patent Holdings, SAS
- Inventors: Gagan Bihari Rath; Karam Naser; Kevin Reuze; Franck Galpin
- Web page: https://patentgrep.com/patent/US20260101033A1

## CPC

- H04N 19/119: Adaptive subdivision aspects, e.g. subdivision of a picture into rectangular or non-rectangular coding blocks
- H04N 19/11: among a plurality of spatial predictive coding modes
- H04N 19/12: Selection from among a plurality of transforms or standards, e.g. selection between discrete cosine transform [DCT] and sub-band transform or selection between H.263 and H.264
- H04N 19/122: Selection of transform size, e.g. 8x8 or 2x4x8 DCT; Selection of sub-band transforms of varying structure or type
- H04N 19/159: Prediction type, e.g. intra-frame, inter-frame or bidirectional frame prediction
- H04N 19/176: the region being a block, e.g. a macroblock
- H04N 19/593: involving spatial prediction techniques
- H04N 19/70: characterised by syntax aspects related to video coding, e.g. related to compression standards
- H04N 19/96: Tree coding, e.g. quad-tree coding
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

An encoding method of an image block is disclosed. The image block to be encoded is first partitioned in at least two partitions, wherein one of said at least two partitions is an L-shaped partition. Each partition is then encoded in encoded data.

## Claims

### Claim 1

An encoding method comprising:
partitioning a current block to be encoded using intra prediction sub-partition (ISP) mode into at least two partitions, wherein one of the at least two partitions is a L-shaped partition; and
encoding the current block by encoding each of the at least two partitions with an intra prediction mode associated with the current block.

### Claim at position 2

2-6. (canceled)

### Claim 7 (depends on 1)

The method of claim 1, wherein predicting the L-shaped partition is performed over the L-shaped partition only.

### Claim 8 (depends on 1)

The method of claim 1, comprising encoding at least one syntax element indicating a configuration for the L-shaped partition among top-left, top-right, bottom-left and bottom-right configurations.

### Claim 9 (depends on 1)

The method of claim 1, wherein the current block to be encoded is partitioned in three partitions, wherein two of the three partitions are L-shaped partitions.

### Claim 10 (depends on 1)

The method of claim 1, wherein encoding the L-shaped partition comprises transforming prediction residuals into transform coefficients, wherein transforming the prediction residuals into transform coefficients comprises:
splitting, in an horizontal direction, the L-shaped partition into a first rectangular block and a second rectangular block, the first rectangular block being larger than the second rectangular block;
applying a first right transform on the first rectangular block to obtain a first intermediate block of transform coefficients and applying a second right transform on the second rectangular block to obtain a second intermediate block of transform coefficients the first and second intermediate blocks of transform coefficients forming an L-shape block of transform coefficients;
splitting, in a vertical direction, the L-shaped block of transform coefficients into two rectangular blocks; and
applying one left transform on each of the two rectangular blocks to obtain an L-shaped block of transform coefficients.

### Claim at position 7

11-13. (canceled)

### Claim 14

A decoding method comprising:
obtaining encoded data for a current block to be decoded using intra prediction sub-partition (ISP) mode into at least two partitions, wherein one of the at least two partitions is a L-shaped partition; and
decoding the current block by decoding each of the at least two partitions from the encoded data with an intra prediction mode associated with the current block.

### Claim at position 9

15-19. (canceled)

### Claim 20 (depends on 14)

The method of claim 14, wherein predicting the L-shaped partition is performed over the L-shaped partition only.

### Claim 21 (depends on 14)

The method of claim 14, comprising decoding at least one syntax element indicating a configuration for the L-shaped partition among top-left, top-right, bottom-left and bottom-right configurations.

### Claim 22 (depends on 14)

The method of claim 14, wherein the current block to be decoded is partitioned in three partitions, wherein two of the three partitions are L-shaped partitions.

### Claim 23 (depends on 14)

The method of claim 14, wherein decoding the L-shaped partition comprises inverse transforming a L-shaped block of transform coefficients into prediction residuals, wherein inverse transforming a L-shaped block of transform coefficients into prediction residuals comprises:
splitting, in a vertical direction, the L-shaped block of transform coefficients into a first rectangular block and a second rectangular block, the first rectangular block being larger than the second rectangular block;
applying a first left inverse transform on the first rectangular block to obtain a first intermediate block of transform coefficients and applying a second left inverse transform on the second rectangular block to obtain a second intermediate block of transform coefficients, the first and second intermediate blocks of transform coefficients forming a new L-shape block of transform coefficients;
splitting, in a horizontal direction, the new L-shaped block of transform coefficients into two rectangular blocks of transform coefficients; and
applying one right inverse transform on each of the two rectangular blocks to obtain an L-shaped block of prediction residuals.

### Claim at position 14

24-30. (canceled)

### Claim 31 (depends on 1)

The method of claim 1, wherein lengths of all sides of the L-shaped partition are a power of two.

### Claim 32 (depends on 1)

The method of claim 1, wherein encoding each of the at least two partitions with an intra prediction mode associated with the current block comprises:
predicting the L-shape partition with the intra prediction mode associated with the current block to obtain a L-shape prediction;
encoding and reconstructing the L-shape partition based on the L-shape prediction;
predicting the other partition with the intra prediction mode associated with the current block based on the reconstructed L-shape partition to obtain another partition prediction; and
encoding the other partition based on the another partition prediction.

### Claim 33 (depends on 14)

The method of claim 14, wherein lengths of all sides of the L-shaped partition are a power of two.

### Claim 34 (depends on 14)

The method of claim 14, wherein decoding each of the at least two partitions with an intra prediction mode associated with the current block comprises:
predicting the L-shape partition with the intra prediction mode associated with the current block to obtain a L-shape prediction;
reconstructing the L-shape partition based on the L-shape prediction;
predicting the other partition with the intra prediction mode associated with the current block based on the reconstructed L-shape partition to obtain another partition prediction; and
reconstructing the other partition based on the another partition prediction.

### Claim 35

An encoding apparatus comprising one or more processors and at least one memory coupled to the one or more processors, wherein the one or more processors are configured to perform:
partitioning a current block to be encoded using intra prediction sub-partition (ISP) mode into at least two partitions, wherein one of the at least two partitions is a L-shaped partition; and
encoding the current block by encoding each of the at least two partitions with an intra prediction mode associated with the current block.

### Claim 36 (depends on 35)

The encoding apparatus of claim 35, wherein lengths of all sides of the L-shaped partition are a power of two.

### Claim 37 (depends on 35)

The encoding apparatus of claim 35, wherein encoding each of the at least two partitions with an intra prediction mode associated with the current block comprises:
predicting the L-shape partition with the intra prediction mode associated with the current block to obtain a L-shape prediction;
encoding and reconstructing the L-shape partition based on the L-shape prediction;
predicting the other partition with the intra prediction mode associated with the current block based on the reconstructed L-shape partition to obtain another partition prediction; and
encoding the other partition based on the another partition prediction.

### Claim 38

A decoding apparatus comprising one or more processors and at least one memory coupled to the one or more processors, wherein the one or more processors are configured to perform:
obtaining encoded data for a current block to be decoded using intra prediction sub-partition (ISP) mode into at least two partitions, wherein one of the at least two partitions is a L-shaped partition; and
decoding the current block by decoding each of the at least two partitions from the encoded data with an intra prediction mode associated with the current block.

### Claim 39 (depends on 38)

The decoding apparatus of claim 38, wherein lengths of all sides of the L-shaped partition are a power of two.

### Claim 40 (depends on 38)

The decoding apparatus of claim 38, wherein decoding each of the at least two partitions with an intra prediction mode associated with the current block comprises:
predicting the L-shape partition with the intra prediction mode associated with the current block to obtain a L-shape prediction;
reconstructing the L-shape partition based on the L-shape prediction;
predicting the other partition with the intra prediction mode associated with the current block based on the reconstructed L-shape partition to obtain another partition prediction; and
reconstructing the other partition based on the another partition prediction.
