Point Cloud Entropy Decoding with Adaptive Probability Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current entropy coding methods for point cloud compression exhibit high encoding and decoding complexity, which hampers the performance of geometry-based point cloud compression (G-PCC).
Innovation Solution
A method for point cloud encoding and decoding that involves determining and signaling flag information corresponding to sequence and geometry parameter sets to adaptively select a probability model, allowing for entropy context continuation and simplifying encoding and decoding operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If entropy coding methods are used for point cloud compression, then coding efficiency is improved, but encoding and decoding complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-defining multiple probability models and their selection conditions before the actual encoding/decoding process. The encoder and decoder are configured with a plurality of probability models and selection conditions in advance, allowing them to quickly select appropriate models during coding without performing complex real-time analysis, thus maintaining high coding efficiency while reducing operational complexity
Solution Approach 2:
The patent changes parameters by switching between different probability models based on local context characteristics. Instead of using a single complex adaptive model, the system selects from multiple predefined probability models (with different parameter configurations) according to simple selection conditions, achieving adaptive coding performance with reduced computational complexity
2Productivity
If adaptive probability model selection based on local context is implemented, then coding performance is improved, but computational complexity increases
Solution Approach 1:
The patent prepares multiple probability models with different characteristics before coding begins. Each model is pre-configured for specific local context scenarios, and simple selection conditions are established in advance to determine when to use each model, eliminating the need for complex real-time probability model optimization
Solution Approach 2:
The system achieves adaptive coding by changing which probability model is active based on simple context evaluation. The selection conditions monitor local context characteristics and switch between predefined probability models, providing adaptive performance without the computational burden of continuously optimizing model parameters
Data Source
AI summary
A decoder decodes a bitstream to determine first flag information corresponding to a current processing unit, where the first flag information is flag information corresponding to a sequence parameter set of a point cloud. The decoder further determines second flag information corresponding to the current processing unit according to the first flag information, where the second flag information is flag information corresponding to a geometry parameter set. The decoder further determines a probability model based on the second flag information, and obtains a prediction value of the current processing unit according to the probability model.