Point Cloud Coding Lookup Tables for Faster RAHT Neighbor Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional point cloud coding techniques face inefficiencies, particularly in the neighbor search process of region-adaptive hierarchical transform (RAHT), leading to high time complexity due to frequent binary searches.
Innovation Solution
Implement a lookup table for neighbor search in RAHT, where node indices are stored in a data structure, allowing for faster neighbor location using Morton or Hilbert codes, and employing early termination based on valid neighbor counts to reduce complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If binary search is used for neighbor search in RAHT, then neighbor location can be achieved, but time complexity increases significantly
Solution Approach 1:
The patent pre-computes and stores neighbor node indices in a lookup table during the encoding process. This preliminary action eliminates the need for binary search during actual encoding, significantly reducing coding time while maintaining the same neighbor search functionality.
Solution Approach 2:
The patent creates a lookup table that copies and stores pre-computed neighbor node indices. Instead of performing binary search to find neighbors during encoding, the system retrieves pre-fetched neighbor indices from the lookup table, thereby reducing computational complexity and encoding time.
2Productivity
If lookup table is implemented for neighbor search, then coding time is reduced, but memory requirements increase
Solution Approach 1:
The patent implements a lookup table that stores only the necessary neighbor node indices for each node, rather than storing all possible node information. This local quality approach optimizes memory usage by storing only the specific data needed for neighbor search operations.
Solution Approach 2:
The patent changes the parameter of data storage from storing all node information to storing only neighbor node indices. This parameter change reduces the quantity of data stored in memory while maintaining the functionality of neighbor search, thereby improving coding efficiency without proportionally increasing memory requirements.
Data Source
AI summary
Embodiments of the present disclosure provide a method for point cloud coding. The method comprises: determining, for a conversion between a point cloud sequence comprising at least one point cloud (PC) sample associated with a plurality of nodes and a bitstream of the point cloud sequence, a node index of a first node of the plurality of nodes, wherein the first node is stored in a data structure and is indicated in the data structure by the node index; and performing the conversion based on the position indication and the occupancy information.