3D Point Cloud Encoding Using Prediction Mode Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for encoding three-dimensional data are inefficient in reducing the processing amount during decoding, which hinders the widespread adoption of three-dimensional data applications.
Innovation Solution
A three-dimensional data encoding method that selects a prediction mode based on attribute information of nearby points, calculates a predicted value, computes a prediction residual, and generates a bitstream including the prediction mode, prediction residual, and number of prediction modes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing encoding methods are used for three-dimensional data, then the data can be encoded and stored, but the processing amount during decoding remains excessively large
Solution Approach 1:
The patent applies preliminary action by performing prediction mode selection and residual calculation during the encoding phase. The encoder pre-calculates predicted values using multiple prediction modes, selects the optimal mode, and encodes only the residual differences. This preliminary processing reduces the computational burden during decoding, as the decoder only needs to perform simple residual addition rather than complex prediction calculations.
Solution Approach 2:
The patent segments the attribute information encoding process into distinct components: prediction mode selection, predicted value calculation, residual calculation, and residual encoding. By dividing the attribute data into residuals relative to predicted values, the patent reduces the dynamic range and variability of the data that needs to be transmitted and processed during decoding.
2Measurement precision
If multiple prediction modes are used to calculate predicted values, then the prediction accuracy improves, but the encoding complexity increases
Solution Approach 1:
The patent implements partial action by evaluating multiple prediction modes during encoding but only encoding the residual for the selected optimal mode. The decoder receives information about the number of prediction modes and the selected mode index, allowing it to reproduce the same predicted value without performing exhaustive mode evaluation. This approach achieves high prediction accuracy while limiting the complexity increase to only the encoding phase.
3Measurement precision
If the number of prediction modes is increased, then the predicted value accuracy improves, but the bitstream size increases due to encoding more mode information
Solution Approach 1:
The patent encodes only the necessary information for prediction mode selection: a count of available prediction modes and an index identifying the selected mode. The actual predicted values are not encoded, only their residuals. This partial encoding approach allows the system to support multiple prediction modes for accuracy while keeping the bitstream overhead minimal, as the mode information itself requires very few bits.
Solution Approach 2:
The patent changes the parameter being encoded from absolute attribute values to residual values (differences between actual and predicted values). This transformation reduces the magnitude and variability of the encoded data, leading to more efficient compression. The number of prediction modes and selected mode index are also encoded as compact parameters rather than full prediction data.
Data Source
AI summary
A three-dimensional data encoding method encodes a plurality of three-dimensional points, and includes: selecting one prediction mode from two or more prediction modes for calculating a predicted value of an attribute information item of a first three-dimensional point, in accordance with attribute information items of one or more second three-dimensional points in vicinity of the first three-dimensional point; calculating the predicted value by the selected prediction mode; calculating, as a prediction residual, a difference between a value of the attribute information item of the first three-dimensional point and the calculated predicted value; and generating a first bit stream that includes the selected prediction mode, the prediction residual, and a number of the two or more prediction modes.