IP Library Granted Patent US 12684166
Granted Patent B2
US 12684166 · App. 18/512,426 · Granted Jul 14, 2026

Bifurcated transmission of 3D data encoded as octrees

Inventors: Muhammad Haseeb (New York, NY); Eugene Chai (Murray Hill, NJ); Matteo Varvello (Holmdel, NJ)
Assignee: Nokia Solutions and Networks Oy
H04N19/597G06T17/00H04N19/96
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 12684166
App. No.
18/512,426
Granted
Jul 14, 2026
Kind
B2
Abstract

In certain embodiments, higher-level octree telepresence data for a point cloud of a 3D object is transmitted using a reliable transport mechanism that ensures successful transmission, e.g., via re-transmission of dropped packets, and lower-level octree data for the point cloud is transmitted using a non-reliable transport mechanism that does not guarantee successful transmission. When the lower-level data is contained in one or more of the bottom-most layers in the octree representation, at least some of the point cloud might be able to be reconstructed using successfully received data even when some lower-level data is lost and not re-transmitted. Using non-reliable transmission can reduce latency by avoiding time-consuming re-transmissions. The resulting reconstructed point cloud will be incomplete, but it might be sufficiently complete to provide satisfactory user experience with lower latency that using reliable transport for the entire octree.

Claims (48)

1 . A machine-implemented method for transmitting a multi-level tree representation of a three-dimensional (3D) telepresence point cloud for a 3D object, the method comprising:

a transmitter generating the tree representation of the 3D telepresence point cloud for the 3D object;

the transmitter transmitting higher-level, reliable 3D telepresence data of the tree representation to a receiver using a reliable transport mechanism, wherein lost higher-level, reliable 3D telepresence data needs to be retransmitted; and

the transmitter transmitting lower-level, unreliable 3D telepresence data of the tree representation to the receiver using a non-reliable transport mechanism, wherein:

lost lower-level, unreliable 3D telepresence data does not need to be retransmitted;

the lower-level, unreliable 3D telepresence data corresponds to two or more bottom levels in the tree representation;

a number of nodes in each of the two or more bottom levels is equal to a number of points in the 3D telepresence point cloud; and

the nodes in the two or more bottom levels have a one-to-one correspondence.

2 . The method of claim 1 , wherein the lower-level data is reordered lower-level data to spread burst noise over the point cloud.

3 . The method of claim 1 , wherein each level in the tree representation has a unique sequence of possible index values for the level's occupancy data.

4 . The method of claim 1 , wherein the tree representation is an octree representation.

5 . The method of claim 1 , wherein the transmitter strategically discards some of the 3D telepresence data.

6 . A transmitter for transmitting a multi-level tree representation of a three-dimensional (3D) telepresence point cloud for a 3D object, the transmitter comprising

at least one processor; and

at least one memory storing instructions that, upon being executed by the at least one processor, cause the transmitter at least to:

generate the tree representation of the 3D telepresence point cloud for the 3D object;

transmit higher-level, reliable 3D telepresence data of the tree representation to a receiver using a reliable transport mechanism, wherein lost higher-level, reliable 3D telepresence data needs to be retransmitted; and

transmit lower-level, unreliable 3D telepresence data of the tree representation to the receiver using a non-reliable transport mechanism, wherein:

lost lower-level, unreliable 3D telepresence data does not need to be retransmitted;

the lower-level, unreliable 3D telepresence data corresponds to two or more bottom levels in the tree representation;

a number of nodes in each of the two or more bottom levels is equal to a number of points in the 3D telepresence point cloud; and

the nodes in the two or more bottom levels have a one-to-one correspondence.

7 . The transmitter of claim 6 , wherein the lower-level data is reordered lower-level data to spread burst noise over the point cloud.

8 . The transmitter of claim 6 , wherein each level in the tree representation has a unique sequence of possible index values for the level's occupancy data.

9 . The transmitter of claim 6 , wherein the tree representation is an octree representation.

10 . The transmitter of claim 6 , wherein the transmitter is configured to strategically discard some of the 3D telepresence data.

11 . A machine-implemented method for receiving a multi-level tree representation of a three-dimensional (3D) telepresence point cloud for a 3D object, the method comprising:

a receiver receiving higher-level, reliable 3D telepresence data of the tree representation from a transmitter via a reliable transport mechanism, wherein lost higher-level, reliable 3D telepresence data needs to be retransmitted;

the receiver receiving lower-level, unreliable 3D telepresence data of the tree representation from the transmitter via a non-reliable transport mechanism, wherein lost lower-level, unreliable 3D telepresence data does not need to be retransmitted; and

the receiver decoding the tree representation to reconstruct the 3D telepresence point cloud for the 3D object, wherein:

the lower-level, unreliable 3D telepresence data corresponds to two or more bottom levels in the tree representation;

a number of nodes in each of the two or more bottom levels is equal to a number of points in the 3D telepresence point cloud; and

the nodes in the two or more bottom levels have a one-to-one correspondence.

12 . The method of claim 11 , wherein the lower-level data is reordered lower-level data to spread burst noise over the point cloud.

13 . The method of claim 11 , wherein each level in the tree representation has a unique sequence of possible index values for the level's occupancy data.

14 . The method of claim 10 , wherein the tree representation is an octree representation.

15 . A receiver for receiving a multi-level tree representation of a three-dimensional (3D) telepresence point cloud for a 3D object, the receiver comprising:

at least one processor; and

at least one memory storing instructions that, upon being executed by the at least one processor, cause the receiver at least to:

receive higher-level, reliable 3D telepresence data of the tree representation from a transmitter via a reliable transport mechanism, wherein lost higher-level, reliable 3D telepresence data needs to be retransmitted;

receive lower-level, unreliable 3D telepresence data of the tree representation from the transmitter via a non-reliable transport mechanism, wherein lost lower-level, unreliable 3D telepresence data does not need to be retransmitted; and

decode the tree representation to reconstruct the 3D telepresence point cloud for the 3D object, wherein:

the lower-level, unreliable 3D telepresence data corresponds to two or more bottom levels in the tree representation;

a number of nodes in each of the two or more bottom levels is equal to a number of points in the 3D telepresence point cloud; and

the nodes in the two or more bottom levels have a one-to-one correspondence.

16 . The receiver of claim 15 , wherein the lower-level data is reordered lower-level data to spread burst noise over the point cloud.

17 . The receiver of claim 15 , wherein each level in the tree representation has a unique sequence of possible index values for the level's occupancy data.

18 . The receiver of claim 15 , wherein the tree representation is an octree representation.