Intra prediction method and decoder for geometric-based point cloud coding
An intra prediction method and a decoder are provided. The method includes the following. Coordinate information of a current node and a current level corresponding to the current node are determined. A current quantity is determined based on the current level, where the current quantity is positively correlated with the current level. Current neighbour nodes are determined based on the coordinate information of the current node, where the current neighbour nodes are neighbour nodes of the current quantity among neighbour nodes of a parent node of the current node. Intra prediction is performed on the current node based on the coordinate information of the current node and the current neighbour nodes.
1. An intra prediction method, for an encoder and comprising:
determining coordinate information of a current node and a current level corresponding to the current node;
determining a current quantity of current neighbour nodes used in intra prediction on the current node based on the current level, wherein the current quantity is positively correlated with the current level, and the current neighbour nodes belongs to neighbour nodes of a parent node of the current node;
determining the current neighbour nodes based on the coordinate information of the current node; and
performing intra prediction on the current node based on the coordinate information of the current node and the current neighbour nodes.
2. The method of claim 1 , wherein determining the current quantity based on the current level, and determining the current neighbour nodes based on the coordinate information of the current node comprises:
determining the current quantity to be a first preset quantity when the current level is greater than or equal to a preset prediction level and less than or equal to a preset level;
determining first neighbour nodes of the first preset quantity adjacent to the current node among the neighbour nodes of the parent node of the current node, according to a preset shift parameter and the coordinate information of the current node;
determining the current quantity to be a second preset quantity when the current level is greater than the preset level; and
determining second neighbour nodes of the second preset quantity among the neighbour nodes of the parent node of the current node, according to the preset shift parameter and the coordinate information of the current node, wherein
the first preset quantity is less than the second preset quantity, and the current neighbour nodes are the first neighbour nodes or the second neighbour nodes.
3. The method of claim 1 , wherein determining the current quantity based on the current level, and determining the current neighbour nodes based on the coordinate information of the current node comprises:
determining a density of points in each neighbour node of the parent node of the current node based on the current level and the coordinate information of the current node when the current level is greater than or equal to a preset prediction level; and
determining the current neighbour nodes based on the density of the points, wherein
the density of the points is negatively correlated with the current quantity.
4. The method of claim 1 , wherein determining the current neighbour nodes based on the coordinate information of the current node comprises:
determining coordinate information and occupancy information of the current neighbour nodes based on a preset shift parameter and the coordinate information of the current node.
5. The method of claim 4 , wherein performing intra prediction on the current node based on the coordinate information of the current node and the current neighbour nodes comprises:
determining node distances between the current node and the current neighbour nodes based on the coordinate information of the current node and the coordinate information of the current neighbour nodes;
determining, from a preset weight table, weights of the current neighbour nodes based on the node distances and the occupancy information;
determining an average weight of the current neighbour nodes based on the weights of the current neighbour nodes; and
performing intra prediction on the current node based on the average weight and determining an intra prediction result.
6. The method of claim 5 , wherein performing intra prediction on the current node based on the average weight and determining the intra prediction result comprises:
performing intra prediction on the current node and determining that the intra prediction result is “occupied”, when the average weight is greater than an upper weight threshold;
performing intra prediction on the current node and determining that the intra prediction result is “non-occupied”, when the average weight is less than a lower weight threshold; and
determining that the intra prediction result is “not predicted” when the average weight is less than or equal to the upper weight threshold and greater than or equal to the lower weight threshold, wherein
the upper weight threshold comprises a first upper weight threshold and a second upper weight threshold, the lower weight threshold comprises a first lower weight threshold and a second lower weight threshold, the first upper weight threshold and the first lower weight threshold correspond to first neighbour nodes, and the second upper weight threshold and the second lower weight threshold correspond to second neighbour nodes.
7. An intra prediction method, for a decoder and comprising:
determining coordinate information of a current node and a current level corresponding to the current node;
determining a current quantity of current neighbour nodes used in intra prediction on the current node based on the current level, wherein the current quantity is positively correlated with the current level, and the current neighbour nodes belongs to neighbour nodes of a parent node of the current node;
determining the current neighbour nodes based on the coordinate information of the current node; and
performing intra prediction on the current node based on the coordinate information of the current node and the current neighbour nodes.
8. The method of claim 7 , wherein determining the current quantity based on the current level, and determining the current neighbour nodes based on the coordinate information of the current node comprises:
determining the current quantity to be a first preset quantity when the current level is greater than or equal to a preset prediction level and less than or equal to a preset level;
determining first neighbour nodes of the first preset quantity adjacent to the current node among the neighbour nodes of the parent node of the current node, according to a preset shift parameter and the coordinate information of the current node;
determining the current quantity to be a second preset quantity when the current level is greater than the preset level; and
determining second neighbour nodes of the second preset quantity among the neighbour nodes of the parent node of the current node, according to the preset shift parameter and the coordinate information of the current node, wherein
the first preset quantity is less than the second preset quantity, and the current neighbour nodes are the first neighbour nodes or the second neighbour nodes.
9. The method of claim 7 , wherein determining the current quantity based on the current level, and determining the current neighbour nodes based on the coordinate information of the current node comprises:
determining a density of points in each neighbour node of the parent node of the current node based on the current level and the coordinate information of the current node when the current level is greater than or equal to a preset prediction level; and
determining the current neighbour nodes based on the density of the points, wherein
the density of the points is negatively correlated with the current quantity.
10. The method of claim 7 , wherein determining the current neighbour nodes based on the coordinate information of the current node comprises:
determining coordinate information and occupancy information of the current neighbour nodes based on a preset shift parameter and the coordinate information of the current node.
11. The method of claim 10 , wherein performing intra prediction on the current node based on the coordinate information of the current node and the current neighbour nodes comprises:
determining node distances between the current node and the current neighbour nodes based on the coordinate information of the current node and the coordinate information of the current neighbour nodes;
determining, from a preset weight table, weights of the current neighbour nodes based on the node distances and the occupancy information;
determining an average weight of the current neighbour nodes based on the weights of the current neighbour nodes; and
performing intra prediction on the current node based on the average weight and determining an intra prediction result.
12. The method of claim 11 , wherein performing intra prediction on the current node based on the average weight and determining the intra prediction result comprises:
performing intra prediction on the current node and determining that the intra prediction result is “occupied”, when the average weight is greater than an upper weight threshold;
performing intra prediction on the current node and determining that the intra prediction result is “non-occupied”, when the average weight is less than a lower weight threshold; and
determining that the intra prediction result is “not predicted” when the average weight is less than or equal to the upper weight threshold and greater than or equal to the lower weight threshold, wherein
the upper weight threshold comprises a first upper weight threshold and a second upper weight threshold, the lower weight threshold comprises a first lower weight threshold and a second lower weight threshold, the first upper weight threshold and the first lower weight threshold correspond to first neighbour nodes, and the second upper weight threshold and the second lower weight threshold correspond to second neighbour nodes.
13. A decoder comprising:
at least one processor; and
a memory coupled to the at least one processor and storing at least one computer executable instruction thereon which, when executed by the at least one processor, causes the at least one processor to:
determine coordinate information of a current node and a current level corresponding to the current node;
determine a current quantity of current neighbour nodes used in intra prediction on the current node based on the current level, wherein the current quantity is positively correlated with the current level, and the current neighbour nodes belongs to neighbour nodes of a parent node of the current node;
determine the current neighbour nodes based on the coordinate information of the current node; and
perform intra prediction on the current node based on the coordinate information of the current node and the current neighbour nodes.
14. The decoder of claim 13 , wherein the at least one processor configured to determine the current quantity based on the current level, and determine the current neighbour nodes based on the coordinate information of the current node is configured to:
determine the current quantity to be a first preset quantity when the current level is greater than or equal to a preset prediction level and less than or equal to a preset level;
determine first neighbour nodes of the first preset quantity adjacent to the current node among the neighbour nodes of the parent node of the current node, according to a preset shift parameter and the coordinate information of the current node;
determine the current quantity to be a second preset quantity when the current level is greater than the preset level; and
determine second neighbour nodes of the second preset quantity among the neighbour nodes of the parent node of the current node, according to the preset shift parameter and the coordinate information of the current node, wherein
the first preset quantity is less than the second preset quantity, and the current neighbour nodes are the first neighbour nodes or the second neighbour nodes.
15. The decoder of claim 13 , wherein the at least one processor configured to determine the current quantity based on the current level, and determine the current neighbour nodes based on the coordinate information of the current node is configured to:
determine a density of points in each neighbour node of the parent node of the current node based on the current level and the coordinate information of the current node when the current level is greater than or equal to a preset prediction level; and
determine the current neighbour nodes based on the density of the points, wherein
the density of the points is negatively correlated with the current quantity.
16. The decoder of claim 13 , wherein the at least one processor configured to determine the current neighbour nodes based on the coordinate information of the current node is configured to:
determine coordinate information and occupancy information of the current neighbour nodes based on a preset shift parameter and the coordinate information of the current node.
17. The decoder of claim 16 , wherein the at least one processor configured to perform intra prediction on the current node based on the coordinate information of the current node and the current neighbour nodes is configured to:
determine node distances between the current node and the current neighbour nodes based on the coordinate information of the current node and the coordinate information of the current neighbour nodes;
determine, from a preset weight table, weights of the current neighbour nodes based on the node distances and the occupancy information;
determine an average weight of the current neighbour nodes based on the weights of the current neighbour nodes; and
perform intra prediction on the current node based on the average weight and determine an intra prediction result.
18. The decoder of claim 17 , wherein the at least one processor configured to perform intra prediction on the current node based on the average weight and determine the intra prediction result is configured to:
perform intra prediction on the current node and determine that the intra prediction result is “occupied”, when the average weight is greater than an upper weight threshold;
perform intra prediction on the current node and determine that the intra prediction result is “non-occupied”, when the average weight is less than a lower weight threshold; and
determine that the intra prediction result is “not predicted” when the average weight is less than or equal to the upper weight threshold and greater than or equal to the lower weight threshold, wherein
the upper weight threshold comprises a first upper weight threshold and a second upper weight threshold, the lower weight threshold comprises a first lower weight threshold and a second lower weight threshold, the first upper weight threshold and the first lower weight threshold correspond to first neighbour nodes, and the second upper weight threshold and the second lower weight threshold correspond to second neighbour nodes.
19. The decoder of claim 14 , wherein the first preset quantity is 6 and the second preset quantity is 26.
20. An encoder, comprising at least one processor and a memory coupled to the at least one processor and storing at least one computer executable instruction thereon which, when executed by the at least one processor, causes the at least one processor to perform the method of claim 1 .