Predictive tree coding for point cloud coding

1. A method for decoding a point cloud using octree partitioning and a predictive tree by at least one processor, the method comprising:

obtaining the point cloud;

determining a bounding box of the point cloud;

generating octree nodes by partitioning the bounding box using octree partitioning;

determining whether to enable the predictive tree based on determining whether an octree partition depth of the bounding box satisfies a threshold by comparing the octree partition depth to the threshold;

based on determining that the octree partition depth of the bounding box satisfies the threshold by comparing the octree partition depth to the threshold, generating the predictive tree for points in at least one octree node of the octree nodes;

applying a transform to the predictive tree; and

decoding the points in the at least one octree node using the predictive tree.

2. The method of claim 1 , further comprising:

identifying a flag that specifies that the points in the at least one octree node are coded using predictive tree coding; and

generating the predictive tree for the points in the at least one octree node, based on identifying the flag.

3. The method of claim 1 , wherein the threshold is predefined.

4. The method of claim 1 , wherein the threshold is signaled in a high-level syntax.

5. The method of claim 1 , further comprising:

generating the predictive tree using geometry information and attribute information of the points.

6. The method of claim 1 , further comprising:

preventing octree partitioning of the at least one node, based on generating the predictive tree.

7. The method of claim 1 , wherein the transform is a discrete cosine transform (DCT).

8. The method of claim 1 , wherein the transform is a discrete sine transform (DST).

9. The method of claim 1 , wherein the transform is a discrete wavelet transform (DWT).

10. The method of claim 1 , wherein the transform is a graph-based transform.

11. The method of claim 1 , wherein the transform is a region adaptive hierarchical transform (RAHT).

12. A device for decoding a point cloud using octree partitioning and a predictive tree, the device comprising:

at least one memory configured to store program code; and

at least one processor configured to read the program code and operate as instructed by the program code, the program code including:

obtaining code configured to cause the at least one processor to obtain the point cloud;

first determining code configured to cause the at least one processor to determine a bounding box of the point cloud;

first generating code configured to cause the at least one processor to generate octree nodes by partitioning the bounding box using octree partitioning;

second determining code configured to cause the at least one processor to determine whether to enable the predictive tree based on determining whether an octree partition depth of the bounding box satisfies a threshold by comparing the octree partition depth to the threshold;

second generating code configured to cause the at least one processor to, based on determining that the octree partition depth of the bounding box satisfies the threshold by comparing the octree partition depth to the threshold, generate the predictive tree for points in at least one octree node of the octree nodes

applying code configured to cause the at least one processor to apply a transform to the predictive tree; and

decoding code configured to cause the at least one processor to decode the points in the at least one octree node using the predictive tree.

13. The device of claim 12 , further comprising:

identifying code configured to cause the at least one processor to identify a flag that specifies that the points in the at least one octree node are coded using predictive tree coding,

wherein the generating code is configured to cause the at least one processor to generate the predictive tree for the points in the at least one octree node, based on identifying the flag.

14. The device of claim 12 , wherein the threshold is predefined.

15. The device of claim 12 , wherein the threshold is signaled in a high-level syntax.

16. The device of claim 12 , wherein the generating code is configured to cause the at least one processor generate the predictive tree using geometry information and attribute information of the points.

17. The device of claim 12 , wherein the transform is at least one of a discrete cosine transform (DCT), a discrete sine transform (DST), a discrete wavelet transform (DWT), a graph-based transform, or a region adaptive hierarchical transform (RAHT).

18. A non-transitory computer-readable medium storing instructions, the instructions comprising: one or more instructions that, when executed by one or more processors of a device for decoding a point cloud using octree partitioning and a predictive tree, cause the one or more processors to:

obtain the point cloud;

determine a bounding box of the point cloud;

generate octree nodes by partitioning the bounding box using octree partitioning;

determine whether to enable the predictive tree based on determining whether an octree partition depth of the bounding box satisfies a threshold by comparing the octree partition depth to the threshold;

based on determining that the octree partition depth of the bounding box satisfies the threshold by comparing the octree partition depth to the threshold, generate the predictive tree for points in at least one octree node of the octree nodes;

apply a transform to the predictive tree; and

decode the points in the at least one octree node using the predictive tree.