Point Cloud Attribute Prediction Using Morton Code Neighbor Table
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In Point Cloud Exploration Model (PCEM)-based coder frameworks, predicting attribute information using Morton codes is inaccurate due to periodical jump points, leading to significant prediction residuals and reduced coding efficiency.
Innovation Solution
A method for predicting attribute information that involves determining a current Morton code for a point in a point cloud, using a preset neighbor information table to find a target Morton code based on coordinate differences, and then judging the existence of a neighbor point to perform prediction using attribute reconstruction information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If Morton codes are used for attribute information prediction, then the coding framework is simple and easy to implement, but the prediction accuracy is low due to periodical jump points
Solution Approach 1:
The patent segments the prediction process into multiple stages: first using Morton codes for initial ordering, then identifying periodical jump points, and finally applying alternative prediction methods for affected points. This segmentation allows the system to maintain the simplicity of Morton code-based ordering while correcting its accuracy deficiencies through targeted interventions.
Solution Approach 2:
The patent introduces an intermediary mechanism (the periodical jump point detection and correction system) that mediates between the simple Morton code ordering and the accurate attribute prediction. This intermediary identifies problematic jump points and applies corrective prediction methods, thereby bridging the gap between ease of implementation and prediction accuracy.
2Ease of manufacture
If backward differential method is used based on Morton code, then the coding process is straightforward, but prediction residual is large affecting coding efficiency
Solution Approach 1:
The patent applies preliminary action by pre-identifying periodical jump points in the Morton code sequence before performing attribute prediction. This allows the system to prepare alternative prediction strategies in advance for points that would otherwise suffer from large residuals, thereby improving coding efficiency without complicating the overall coding process.
Solution Approach 2:
The patent changes the prediction parameter dynamically: for most points, it uses the standard backward differential method with Morton code ordering, but for points identified as affected by periodical jump points, it switches to alternative prediction methods. This parameter change optimizes prediction accuracy and reduces residuals while maintaining coding efficiency.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The embodiments of the present disclosure provide a method for predicting attribute information, a coder, a decoder, and a storage medium. The coder determines a current Morton code corresponding to a point to be predicted in a point cloud to be coded, determines a target Morton code corresponding to the point to be predicted based on the current Morton code and according to a preset neighbor information table, judges whether a neighbor point of the point to be predicted exists in the point cloud to be coded according to the target Morton code, and performs prediction to obtain a predicted attribute value of the point to be predicted according to attribute reconstruction information of the neighbor point in response to that the neighbor point exists in the point cloud to be coded. The decoder determines a current Morton code corresponding to a point to be predicted in a point cloud to be decoded, determines a target Morton code corresponding to the point to be predicted based on the current Morton code and according to a preset neighbor information table, judges whether a neighbor point of the point to be predicted exists in the point cloud to be decoded according to the target Morton code, and performs prediction to obtain a predicted attribute value of the point to be predicted according to attribute reconstruction information of the neighbor point in response to that the neighbor point exists in the point cloud to be decoded.