3D Point Cloud Encoding With Prediction Tree Bit-Count Compression

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing three-dimensional data encoding and decoding methods lack efficient coding techniques, leading to inefficient data compression and transmission of massive point cloud data.

Innovation Solution

The proposed method encodes geometry information using a prediction tree and represents the child node count information with a bit count, optimizing the encoding process by using a lowest bit count for value 1 and potentially higher bits for other values.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional encoding methods are used for three-dimensional point cloud data, then the encoding process is simple, but the coding efficiency is low and data compression is insufficient

Engineering Contradiction:

Improvecoding efficiencyVSAvoidencoding process complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The encoding process is divided into distinct stages: geometry encoding using prediction trees, child node count encoding with variable bit counts, and attribute encoding. This segmentation allows each component to be optimized independently, improving overall coding efficiency while maintaining manageable complexity through modular design.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies parameter changes by using variable bit counts for encoding child node counts based on the actual number of child nodes. When a node has fewer child nodes, fewer bits are used, dynamically adapting the encoding parameters to the data characteristics and improving compression efficiency.

Inventive Principle:

Principle #35Parameter changes

2Quantity of substance

If variable bit counts are used for child node count encoding, then data compression is improved, but the decoding complexity increases

Engineering Contradiction:

Improvedata sizeVSAvoiddecoding process complexity

Core Design Contradiction:

Quantity of substanceVSDevice complexity

Solution Approach 1:

The encoding process performs preliminary organization of data into prediction tree structures with explicit child node count information. This preliminary action during encoding simplifies the decoding process, as the decoder can directly read the pre-encoded child node counts without performing complex analysis, reducing decoding complexity while maintaining compression efficiency.

Inventive Principle:

Principle #10Preliminary action

3Productivity

If prediction trees are used for geometry encoding, then encoding efficiency is improved, but the memory requirements increase

Engineering Contradiction:

Improveencoding efficiencyVSAvoidmemory usage

Core Design Contradiction:

ProductivityVSQuantity of substance

Solution Approach 1:

The prediction tree structure implements a nested organization where parent nodes contain references to child nodes, and child nodes are organized in hierarchical levels. This nesting allows efficient traversal and processing of geometry data, improving encoding efficiency while using memory only for the tree structure itself rather than storing redundant geometry information at each level.

Inventive Principle:

Principle #7Nested doll (Nesting)

Data Source

PatentUS12536707B2Three-dimensional data encoding method, three-dimensional data decoding method, three-dimensional data encoding device, and three-dimensional data decoding device

Publication Date: 2026.01.27 PANASONIC INTELLECTUAL PROPERTY CORP OF AMERICA

AI summary

A three-dimensional data encoding method includes: encoding geometry information of a three-dimensional point included in point cloud data, using a prediction tree, to generate encoded geometry information; and encoding information indicating a child node count, which is a total number of child nodes of a node included in the prediction tree, to generate encoded child node count information. The encoded child node count information is represented by a bit count corresponding to the child node count, and when the child node count is 1, the encoded child node count information is represented by a lowest bit count.