# US 2024/0205431 A1 · Information processing apparatus and method

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

- Number: US 2024/0205431 A1
- Document ID: US20240205431A1
- Status: Published application
- Kind: A1
- Application number: 18/588,025
- Filed: 2024-02-27
- Published: 2024-06-20
- Earliest claimed: 2019-03-19
- Assignees: Sony Group Corporation
- Applicants: Sony Group Corporation
- Inventors: Tsuyoshi KATO; Satoru KUMA; Ohji NAKAGAMI; Hiroyuki YASUDA; Koji YANO
- Web page: https://patentgrep.com/patent/US20240205431A1

## CPC

- G06T 17/005: {Tree description, e.g. octree, quadtree}
- H04N 19/30: using hierarchical techniques, e.g. scalability (H04N19/63 takes precedence)
- G06T 9/001: {Model-based coding, e.g. wire frame}
- G06T 9/004: {Predictors, e.g. intraframe, interframe coding}
- H04N 19/124: Quantisation
- H04N 19/1883: {the unit relating to sub-band structure, e.g. hierarchical level, directional tree, e.g. low-high [LH], high-low [HL], high-high [HH]}
- H04N 19/20: using video object coding
- H04N 19/50: using predictive coding (H04N19/61 takes precedence)
- H04N 19/597: specially adapted for multi-view video sequence encoding
- H04N 19/85: using pre-processing or post-processing specially adapted for video compression
- G06T 2210/56: Particle system, point based geometry or rendering
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

There is provided an information processing apparatus, an image processing apparatus, an encoding device, a decoding device, an information processing method, and a program that facilitate scalable decoding of attribute information. Regarding the attribute information about a point cloud representing a three-dimensional object, a process of classifying the respective points of the point cloud into prediction points that are the points at which difference values between the attribute information and predicted values are left, and reference points that are the points at which the attribute information is referred to during derivation of the predicted values is recursively repeated on the reference points, so that the attribute information is hierarchized. At the time of the hierarchization, the prediction points are selected so that points also exist in voxels of the hierarchical level one level higher than the hierarchical level to which the voxels containing the points of the current hierarchical level belong.

## Claims

### Claim 1

A decoding apparatus comprising:
circuitry configured to perform inverse hierarchization of encoded attribute information of a point cloud representing a three-dimensional object,
wherein the encoded attribute information is information hierarchized with an encoding apparatus by
classifying each of highest-resolution points of the point cloud in a lowest hierarchical level of a tree construction into one of a lowest reference point and a lowest prediction point predicted based on the lowest reference point, and
classifying each of intermediate-resolution points in an intermediate hierarchical level of the tree construction into one of an intermediate reference point and an intermediate prediction point predicted based on the intermediate reference point,
wherein the intermediate hierarchical level is higher than the lowest hierarchical level,
wherein the intermediate-resolution points are less than highest-resolution points,
wherein positional information of the intermediate reference point is substantially the same as positional information of the lowest reference point, and
wherein the inverse hierarchization includes:
deriving a scalable decoding instruction representing a decoding target LOD; and
performing a scalable decoding of the encoded attribute information in accordance with the scalable decoding instruction by deriving a predicted value of the intermediate prediction point using the positional information of the intermediate reference point and attribute information of the intermediate reference point.

### Claim 2 (depends on 1)

The decoding apparatus according to claim 1, wherein
the hierarchized information is derived using at least one of a Lifting transform.

### Claim 3 (depends on 1)

The decoding apparatus according to claim 1, wherein
a quantization weight for the intermediate hierarchical level is greater than a quantization weight for the lowest hierarchical level.

### Claim 4 (depends on 3)

The decoding apparatus according to claim 3, wherein
the quantization weight for the intermediate hierarchical level has no dependency on the quantization weight for the lowest hierarchical level.

### Claim 5

A decoding method comprising:
performing inverse hierarchization of encoded attribute information of a point cloud representing a three-dimensional object,
wherein the encoded attribute information is information hierarchized with an encoding apparatus by
classifying each of highest-resolution points of the point cloud in a lowest hierarchical level of a tree construction into one of a lowest reference point and a lowest prediction point predicted based on the lowest reference point, and
classifying each of intermediate-resolution points in an intermediate hierarchical level of the tree construction into one of an intermediate reference point and an intermediate prediction point predicted based on the intermediate reference point,
wherein the intermediate hierarchical level is higher than the lowest hierarchical level,
wherein the intermediate-resolution points are less than highest-resolution points,
wherein positional information of the intermediate reference point is substantially the same as positional information of the lowest reference point, and
wherein the inverse hierarchization includes:
deriving a scalable decoding instruction representing a decoding target LOD; and
performing a scalable decoding of the encoded attribute information in accordance with the scalable decoding instruction by deriving a predicted value of the intermediate prediction point using the positional information of the intermediate reference point and attribute information of the intermediate reference point.

### Claim 6 (depends on 5)

The decoding method according to claim 5, wherein
the hierarchized information is derived using at least one of a Lifting transform.

### Claim 7 (depends on 5)

The decoding method according to claim 5, wherein
a quantization weight for the intermediate hierarchical level is greater than a quantization weight for the lowest hierarchical level.

### Claim 8 (depends on 7)

The decoding method according to claim 7, wherein
the quantization weight for the intermediate hierarchical level has no dependency on the quantization weight for the lowest hierarchical level.
