Granted Patent B2

US 12,581,114 · App. 17/615,336 · Granted Mar 17, 2026

Inventors: Hyejung Hur (Seoul, KR); Sejin Oh (Seoul, KR)

Assignee: LG Electronics Inc.

H04N19/597G06T9/40H04N19/119H04N19/132H04N19/46H04N19/96

Abstract

Claims (46)

1 . A method of encoding point cloud data, the method comprising:

encoding the point cloud data including geometry information and attribute information, wherein the geometry information represents positions of points of the point cloud data and the attribute information represents attributes of the points of the point cloud data, wherein encoding the point cloud data includes:

encoding the geometry information based on an octree;

encoding the attribute information based on a level of detail (LOD) structure generated by a sampling process; and

transmitting a bitstream including the encoded point cloud data, wherein the bitstream includes attribute parameter set information including:

first information for representing that the attribute information is encoded based on generation of the LOD structure,

second information for representing a sampling type of the LOD structure, and

third information related to a number of points in a sampling range for the octree,

wherein, based on a first value of the sampling type, a point that is closest to a centroid of the points in a node is selected based on a lowest index of the points in the sampling range.

2 . The method of claim 1 ,

wherein the LOD is generated based on morton orders of the points or distance between the points.

3 . A device comprising:

a memory; and

a processor connected to the memory, wherein the processor is configured to:

encode encoding geometry information of point cloud data based on an octree, wherein the geometry information represents positions of points of the point cloud data;

encode attribute information of the point cloud data based on a level of detail (LOD) structure generated by a sampling process, wherein the attribute information represents attributes of the points of the point cloud data; and

transmit a bitstream including the encoded point cloud data, wherein the bitstream includes attribute parameter set information including:

first information for representing that the attribute information is encoded based on generation of the LOD structure,

second information for representing a sampling type of the LOD structure, and

third information related to a number of points in a sampling range for the octree,

wherein, based on a first value of the sampling type, a point that is closest to a centroid of the points in a node is selected based on a lowest index of the points in the sampling range.

4 . The device of claim 3 ,

wherein the LOD is generated based on morton orders of the points or distance between the points.

5 . A method comprising:

obtaining a bitstream including point cloud data, wherein the point cloud data includes geometry information and attribute information, wherein the geometry information represents positions of points of the point cloud data and the attribute information represents attributes of the points of the point cloud data;

decoding the geometry information based on an octree; and

decoding the attribute information based on a level of detail (LOD) structure generated by a sampling process,

wherein the bitstream includes attribute parameter set information including:

first information for representing that the attribute information is encoded based on generation of the LOD structure,

second information for representing a sampling type of the LOD structure, and

third information related to a number of points in a sampling range for the octree,

wherein, based on a first value of the sampling type, a point that is closest to a centroid of the points in a node is selected based on a lowest index of the points in the sampling range.

6 . The method of claim 5 ,

wherein the LOD is generated based on morton orders of the points or distance between the points.

7 . A device comprising:

a memory; and

a processor connected to the memory, wherein the processor is configured to:

obtain a bitstream including point cloud data, wherein the point cloud data includes geometry information and attribute information, and wherein the geometry information represents positions of points of the point cloud data and the attribute information represents attributes of the points of the point cloud data;

decode the geometry information based on an octree; and

decode the attribute information based on a level of detail (LOD) structure generated by a sampling process,

wherein the bitstream includes attribute parameter set information including:

first information for representing that the attribute information is encoded based on generation of the LOD structure,

second information for representing a sampling type of the LOD structure, and

third information related to a number of points in a sampling range for the octree,

wherein, based on a first value of the sampling type, a point that is closest to a centroid of the points in a node is selected based on a lowest index of the points in the sampling range.

8 . The device of claim 7 , wherein the LOD is generated based on morton orders of the points or distance between the points.

Assignments (1)

Continuity (2)

References Cited (14)