3D Point Cloud Encoding Using Normal Vectors to Cut Processing Load

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

There is a demand to reduce the processing amount of three-dimensional data encoding and decoding, as existing methods are inefficient in handling the large amounts of data involved in point cloud compression.

Innovation Solution

The proposed solution involves encoding geometry information and attribute information, including normal vectors, for each three-dimensional point in point cloud data to generate a bitstream, which reduces the processing amount by treating normal vectors as attribute information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If point cloud data is compressed using existing methods, then data transmission and storage are enabled, but the processing amount remains excessively large

Engineering Contradiction:

Improvedata amountVSAvoidprocessing amount

Core Design Contradiction:

Quantity of substanceVSProductivity

Solution Approach 1:

The patent segments point cloud data into multiple data blocks, where each block contains points with similar normal vector orientations. This segmentation enables independent processing of each block, reducing the overall processing amount while maintaining data integrity and compression efficiency.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter representation by encoding normal vectors using directional cosine values relative to reference axes, and further optimizes by representing only two independent components instead of three. This parameter transformation significantly reduces the bit depth required for encoding normal vectors, thereby reducing processing amount.

Inventive Principle:

Principle #35Parameter changes

2Manufacturing precision

If normal vectors are encoded for each three-dimensional point, then geometric accuracy is improved, but code size increases

Engineering Contradiction:

Improvegeometric accuracyVSAvoidcode size

Core Design Contradiction:

Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The patent transforms the parameter representation of normal vectors from three independent components to two independent components by expressing normal vectors in a coordinate system where one component is derived from the other two. This reduces the code size while maintaining geometric accuracy through the use of directional cosines.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent applies different encoding precision to different regions of the point cloud data. By segmenting data blocks based on normal vector orientation similarity, it uses coarser quantization for blocks with similar orientations and finer quantization where needed, optimizing the balance between geometric accuracy and code size.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS20250182337A1Three-dimensional data encoding method, three-dimensional data decoding method, three-dimensional data encoding device, and three-dimensional data decoding device

Publication Date: 2025.06.05 PANASONIC INTELLECTUAL PROPERTY CORP OF AMERICA

AI summary

A three-dimensional data encoding method includes: encoding, for each of three-dimensional points included in point cloud data, geometry information and one or more items of attribute information to generate a bitstream. In the encoding, a normal vector of each of the three-dimensional points is encoded as an item of attribute information included in the one or more items of attribute information of the three-dimensional point.