3D Point Cloud Encoding via Prediction Tree Motion Compensation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current three-dimensional data encoding methods lack efficiency in compressing and transmitting point cloud data, which is essential for applications like autonomous vehicles and infrastructure inspection, due to the absence of effective methods for multiplexing and decoding point cloud data using multiple codecs.

Innovation Solution

A three-dimensional data encoding method that parses syntax to determine prediction trees for inter or intra prediction, and uses motion compensation to improve coding efficiency by referencing encoded points in a second frame, generating a bitstream with motion compensation information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If point cloud data is compressed using traditional encoding methods, then data transmission is enabled, but coding efficiency is insufficient

Engineering Contradiction:

Improvecoding efficiencyVSAvoiddata compression quality

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent segments point cloud data into multiple prediction trees, where each tree represents a specific spatial region or feature group. This segmentation allows independent encoding optimization for different data regions, improving overall coding efficiency while maintaining data integrity through structured organization of the point cloud information

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a hierarchical tree structure dimension to organize point cloud data, transforming the traditional flat data representation into a multi-level hierarchical format. This dimensional change enables more efficient compression by exploiting spatial correlations within each tree while reducing redundancy across different levels of the hierarchy

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Adaptability or versatility

If multiple codecs are used for point cloud data, then multiplexing capability is improved, but system complexity increases

Engineering Contradiction:

Improvemultiplexing capabilityVSAvoidsystem complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent designs a universal prediction tree structure that can accommodate multiple encoding approaches and codecs within a unified framework. This multi-functional structure allows the same basic tree-based organization to support different compression algorithms and prediction methods, enabling multiplexing capability without proportionally increasing system complexity

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The prediction tree acts as an intermediary structure that mediates between raw point cloud data and various encoding/decoding processes. This intermediate representation layer allows multiple codecs to operate on a standardized data structure, simplifying the integration of different coding methods while maintaining their individual advantages

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS12200245B2Three-dimensional data encoding method, three-dimensional data decoding method, three-dimensional data encoding device, and three-dimensional data decoding device

Publication Date: 2025.01.14 PANASONIC INTELLECTUAL PROPERTY CORP OF AMERICA

AI summary

A three-dimensional data encoding method includes: parsing a first syntax to determine, for each of at least one prediction tree in a slice, whether inter prediction or intra prediction is to be performed on the prediction tree, the prediction tree containing at least one three-dimensional point; and parsing a second syntax to determine a motion compensation mode to be performed on the prediction tree when the inter prediction is to be performed on the prediction tree.