IP Library › Granted Patent US 12,034,979
Granted Patent B2
US 12,034,979 · App. 17/605,972 · Granted Jul 9, 2024

Point cloud data transmission device, point cloud data transmission method, point cloud data reception device and point cloud data reception method

Inventors: Hyunmook Oh (Seoul, KR); Sejin Oh (Seoul, KR)
Assignee: LG ELECTRONICS INC.
H04N19/96H04N19/124H04N19/593H04N19/597H04N19/91
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,034,979
App. No.
17/605,972
Granted
Jul 9, 2024
Kind
B2
Abstract

A point cloud data transmission method according to embodiments can comprise the steps of: acquiring point cloud data; encoding geometry information included in the point cloud data; encoding attribute information included in the point cloud data; and transmitting the encoded geometry information and the encoded attribute information.

Claims (68)

1. A method for transmitting point cloud data by an apparatus, the method comprising:

acquiring 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;

encoding the geometry information included in the point cloud data;

encoding the attribute information included in the point cloud data; and

transmitting the encoded geometry information, the encoded attribute information, and signaling information,

wherein the encoding the geometry information comprises:

generating an octree by recursively subdividing a bounding box for the point cloud data, and

entropy encoding the geometry information based on the octree,

wherein the encoding the attribute information comprises:

selecting predicted attribute information of a node to be attribute-encoded based on the octree and an attribute coding method, wherein, when a parent node of the node to be attribute-encoded in the octree includes occupied child nodes, attribute information mapped to one child node of the occupied child nodes is selected for the predicted attribute information,

acquiring residual attribute information based on the attribute information of the node to be attribute-encoded and the predicted attribute information,

quantizing the residual attribute information, and

entropy-encoding the quantized residual attribute information,

wherein the signaling information includes information for indicating a type of the attribute coding method and information related to the quantization.

2. The method of claim 1 , wherein the encoding of the geometry information comprises:

quantizing the geometry information;

generating the octree having an octal tree structure based on the quantized geometry information, wherein the octree includes at least a root node corresponding to a highest layer and a leaf node corresponding to a lowest layer;

approximating a structure of the octree;

reconstructing geometry information based on the generated octree and the approximated octree; and

entropy-encoding an occupancy code of each node of the generated octree or the approximated octree and outputting a geometry bitstream.

3. An apparatus for transmitting point cloud data, the apparatus comprising:

an acquirer configured to acquire 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;

a geometry encoder configured to encode the geometry information included in the point cloud data;

an attribute encoder configured to encode the attribute information included in the point cloud data; and

a transmitter configured to transmit the encoded geometry information, the encoded attribute information, and signaling information,

wherein the geometry encoder comprises:

an octree generator configured to generate an octree by recursively subdividing a bounding box for the point cloud data, and

a first arithmetic encoder configured to entropy encode the geometry information based on the octree,

wherein the attribute encoder comprises:

an attribute predictor configured to select predicted attribute information of a node to be attribute-encoded based on the octree and an attribute coding method, wherein, when a parent node of the node to be attribute-encoded in the octree includes occupied child nodes, attribute information mapped to one child node of the occupied child nodes is selected for the predicted attribute information,

a residual attribute estimator configured to acquire residual attribute information based on the attribute information of the node to be attribute-encoded and the predicted attribute information,

a quantizer configured to quantize the residual attribute information, and

a second arithmetic encoder configured to entropy-encode the quantized residual attribute information,

wherein the signaling information includes information for indicating a type of the attribute coding method and information related to the quantization.

4. The apparatus of claim 3 , wherein the geometry encoder comprises:

a quantizer configured to quantize the geometry information;

an octree analyzer configured to generate the octree having an octal tree structure based on the quantized geometry information, wherein the octree includes at least a root node corresponding to a highest layer and a leaf node corresponding to a lowest layer;

an approximation analyzer configured to approximate a structure of the octree;

a geometry reconstructor configured to reconstruct geometry information based on the generated octree and the approximated octree; and

an arithmetic encoder configured to entropy-encode an occupancy code of each node of the generated octree or the approximated octree and output a geometry bitstream.

5. A method for receiving point cloud data by an apparatus, the method comprising:

receiving the point cloud data and signaling information including encoded geometry information and encoded 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 encoded geometry information included in the point cloud data;

decoding the encoded attribute information included in the point cloud data; and

processing and rendering the decoded geometry information and the decoded attribute information,

wherein the decoding the encoded geometry information comprises:

entropy decoding the encoded geometry information, and

generating an octree based on the entropy decoded geometry information,

wherein the decoding the encoded attribute information comprises:

entropy decoding residual attribute information in the encoded attribute information,

inverse quantizing the entropy-decoded residual attribute information,

selecting predicted attribute information of a node to be attribute-decoded based on the octree and an attribute coding method in the signaling information, wherein, when a parent node of the node to be attribute-decoded in the octree includes occupied child nodes, attribute information mapped to one child node of the occupied child nodes is selected for the predicted attribute information, and

reconstructing attribute information based on the residual attribute information and the predicted attribute information,

wherein the signaling information includes information for indicating a type of the attribute coding method and information related to the inverse quantization.

6. An apparatus for receiving point cloud data, the apparatus comprising:

a receiver configured to receive the point cloud data and signaling information including encoded geometry information and encoded 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;

a geometry decoder configured to decode the encoded geometry information included in the point cloud data;

an attribute decoder configured to decode the encoded attribute information included in the point cloud data; and

a renderer configured to process and render the decoded geometry information and the decoded attribute information,

wherein the geometry decoder comprises:

a first arithmetic decoder configured to decode the encoded geometry information, and

an octree generator configured to generate an octree based on the entropy decoded geometry information,

wherein the attribute decoder comprises:

a second arithmetic decoder configured to entropy decode residual attribute information in the encoded attribute information,

an inverse quantizer configured to inverse quantize the entropy-decoded residual attribute information,

an attribute predictor configured to select predicted attribute information of a node to be attribute-decoded based on the octree and an attribute coding method in the signaling information, wherein, when a parent node of the node to be attribute-decoded in the octree includes occupied child nodes, attribute information mapped to one child node of the occupied child nodes is selected for the predicted attribute information, and

an attribute reconstructor configured to reconstruct attribute information based on the residual attribute information and the predicted attribute information,

wherein the signaling information includes information for indicating a type of the attribute coding method and information related to the inverse quantization.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 22, 2021
From: OH, HYUNMOOK; OH, SEJIN
To: LG ELECTRONICS INC.
Reel/Frame 057883/0780 →
Priority Claims (1)
KR 10-2019-0031846 · Mar 20, 2019 · national
Continuity (1)
Related Publication 20220286713A1 · Sep 8, 2022
Cited By (1)
US 12,568,247