Point cloud data processing apparatus and method
A point cloud data processing method according to embodiments may comprise the steps of: encoding point cloud data including geometry information and attribute information; and transmitting a bitstream including the encoded point cloud data.
1. A method for processing 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 and the attribute information,
wherein the encoding the geometry information includes generating a tree including levels including points of the geometry information, wherein the geometry information is encoded based on the tree,
wherein the encoding the attribute information includes generating at least one LOD (Level of Detail) based on a Morton order,
wherein one point of the at least one LOD is selected based on the Morton order,
wherein the encoding the attribute information further includes predicting the attribute information by searching a nearest neighbor for the attribute data,
wherein the searching the nearest neighbor includes partitioning points for the attribute data and selecting the nearest neighbor based on a distance related to the portioned points; and
transmitting a bitstream including the encoded point cloud data,
wherein the bitstream includes attribute parameter set information including search range information representing a neighbor point search range for searching neighbor points located around a center of a point.
2. The method of claim 1 , wherein encoding the attribute information includes:
generating a neighbor point set of points belonging to each LOD.
3. The method of claim 2 , wherein the bitstream includes neighbor point set generation information,
the neighbor point set generation information includes:
type information representing a centroid point selecting method type for selecting a centroid point to generate the neighbor point set, and search type information representing a type of the neighbor point search range per an LOD and search type information representing a type of a neighbor point search method.
4. The method of claim 3 , wherein the search type information includes at least one of an all-point search method, a method of searching for a point selected from a sampling range as a neighbor point, a method of primarily searching points in a bounding box in each of areas obtained by dividing a neighbor point search range, a method of performing a primary search based on division of an octree-based neighbor point search range and a center position value of each octree node, or a method of performing a primary search based on calculation of a distance between points selected from a sampling range.
5. A method of processing point cloud data, the method comprising:
receiving 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,
wherein the decoding the geometry information includes generating a tree including levels including points of the geometry information, wherein the geometry information is decoded based on the tree; and
decoding the attribute information,
wherein decoding the attribute information includes generating at least one LOD (Level of Detail) based on a Morton order,
wherein one point of the at least one LOD is selected based on the Morton order,
wherein the decoding the attribute information further includes predicting the attribute information by searching a nearest neighbor for the attribute data,
wherein the searching the nearest neighbor includes partitioning points for the attribute data and selecting the nearest neighbor based on a distance related to the portioned points; and
wherein the bitstream includes attribute parameter set information including search range information representing a neighbor point search range for searching neighbor points located around a center of a point.
6. The method of claim 5 , wherein decoding the attribute information includes:
generating a neighbor point set of points belonging to each LOD.
7. The method of claim 6 , wherein the bitstream includes neighbor point set generation information,
the neighbor point set generation information includes:
type information representing a centroid point selecting method type for selecting a centroid point to generate the neighbor point set, and search type information representing a type of the neighbor point search range per an LOD and search type information representing a type of a neighbor point search method.
8. The method of claim 7 , wherein the search type information includes at least one of an all-point search method, a method of searching for a point selected from a sampling range as a neighbor point, a method of primarily searching points in a bounding box in each of areas obtained by dividing a neighbor point search range, a method of performing a primary search based on division of an octree-based neighbor point search range and a center position value of each octree node, or a method of performing a primary search based on calculation of a distance between points selected from a sampling range.
9. A device configured to process point cloud data, the device comprising:
a receiver configured to receive 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; and
at least one decoder configured to decode the geometry information and the attribute information,
wherein the decoding the geometry information includes generating a tree including levels including points of the geometry information, wherein the geometry information is decoded based on the tree,
wherein decoding the attribute information includes generating at least one LOD (Level of Detail) based on a Morton order,
wherein one point of the at least one LOD is selected based on the Morton order,
wherein the decoding the attribute information further includes predicting the attribute information by searching a nearest neighbor for the attribute data,
wherein the searching the nearest neighbor includes partitioning points for the attribute data and selecting the nearest neighbor based on a distance related to the portioned points; and
wherein the bitstream includes attribute parameter set information including search range information representing a neighbor point search range for searching neighbor points located around a center of a point.
10. The device of claim 9 ,
wherein decoding the attribute information includes generating a neighbor point set of points belonging to each LOD.
11. The device of claim 10 , wherein the bitstream includes neighbor point set generation information,
the neighbor point set generation information includes:
type information representing a centroid point selecting method type for selecting a centroid point to generate the neighbor point set, and search type information representing a type of the neighbor point search range per an LOD and search type information representing a type of a neighbor point search method.
12. A device configured to process point cloud data, the device comprising:
an encoder configured to encode 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 and the attribute information,
wherein the encoding the geometry information includes generating a tree including levels including points of the geometry information, wherein the geometry information is encoded based on the tree,
wherein the encoding the attribute information includes generating at least one LOD (Level of Detail) based on a Morton order, and
wherein one point of the at least one LOD is selected based on the Morton order; and
wherein the encoding the attribute information further includes predicting the attribute information by searching a nearest neighbor for the attribute data,
wherein the searching the nearest neighbor includes partitioning points for the attribute data and selecting the nearest neighbor based on a distance related to the portioned points; and
a transmitter configured to transmit a bitstream including the encoded point cloud data,
wherein the bitstream includes attribute parameter set information including search range information representing a neighbor point search range for searching neighbor points located around a center of a point a centroid point.
13. The device of claim 12 ,
wherein encoding the attribute information includes generating a neighbor point set of points belonging to each LOD.
14. The device of claim 13 , the bitstream includes neighbor point set generation information,
the neighbor point set generation information includes:
type information representing a centroid point selecting method type for selecting a centroid point to generate the neighbor point set, and search type information representing a type of the neighbor point search range per an LOD and search type information representing a type of a neighbor point search method.