Predictive Point Cloud Geometry Coding From Planar Node Structure
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current point cloud encoding technologies, such as G-PCC, inefficiently encode planar position information due to insufficient consideration of time domain correlation and prior reference information, leading to reduced geometry coding efficiency.
Innovation Solution
Determine planar structure information of a preset node based on a prediction frame, derive context indication information, and use target context information to encode and decode planar position information, considering the planar structure of neighboring nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If predictive encoding is performed on planar position information using only some prior reference information, then the encoding process is simple, but geometry coding efficiency is reduced
Solution Approach 1:
The patent extends the reference information from spatial dimensions to include temporal dimension by introducing prediction frame nodes. The context information is determined based on planar structure information from prediction frame nodes and target nodes, adding a time dimension to the encoding process. This enables the encoder to utilize both spatial and temporal correlations for more accurate prediction.
Solution Approach 2:
The patent performs preliminary determination of planar structure information from prediction frame nodes before the actual encoding process. By pre-determining the planar structure information and using it to derive context indication information, the system prepares reference data in advance that improves the accuracy of subsequent predictive encoding.
2Productivity
If more comprehensive reference information is used for predictive encoding, then geometry coding efficiency is improved, but the encoding process becomes more complex
Solution Approach 1:
The patent segments the reference information into distinct components: prediction frame nodes and target nodes. Each node has specific planar structure information that is determined separately, and then combined to form the context indication information. This segmentation allows for organized processing of multiple reference sources without overwhelming complexity.
Solution Approach 2:
The patent dynamically determines context information based on the actual planar structure information extracted from prediction frame nodes and target nodes. The context indication information is derived adaptively from the determined planar structure, allowing the encoding process to adjust to different geometric configurations while maintaining efficient use of reference information.
Data Source
AI summary
Disclosed in the embodiments of the present application are an encoding method, a decoding method, a bitstream, an encoder, a decoder and a storage medium. The decoding method comprises: determining, based on a prediction node that is in a prediction frame and corresponds to a current node, planar structure information of a preset node of the current node, wherein the preset node comprises the prediction node; determining, based on the planar structure information of the preset node, context indication information of the current node; determining, based on the context indication information, target context information; and decoding a bitstream based on the target context information, to determine planar position information of the current node.