Point Cloud Metadata Encoding for Low-Bitrate Reconstruction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing technologies face challenges in efficiently compressing dynamic point clouds for distribution while maintaining high quality and reducing bit-rate consumption, particularly in applications like virtual reality and autonomous vehicles, where point clouds are often large and require real-time processing.

Innovation Solution

A two-layer-based encoding structure is employed, comprising a base layer for lossy representation and an enhancement layer for lossless details, using existing video codecs to compress geometry and texture information, along with metadata encoding for patch information and occupancy maps to optimize compression efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of energy

If dynamic point clouds are compressed using conventional methods, then bit-rate consumption is reduced, but reconstruction quality deteriorates

Engineering Contradiction:
Improvebit-rate consumptionVSAvoidreconstruction quality
Core Design Contradiction:
Loss of energyVSManufacturing precision

Solution Approach 1:

The patent divides the point cloud compression into two independent layers: a base layer for lossy compression and an enhancement layer for lossless details. This segmentation allows the base layer to achieve high compression ratios while the enhancement layer preserves critical quality information, resolving the contradiction between bit-rate reduction and reconstruction quality.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The enhancement layer selectively encodes only the most important geometric and attribute information that contributes to reconstruction quality. By applying local quality principles, the system maintains high quality where needed while using minimal bits, thus reducing overall bit-rate consumption without sacrificing reconstruction accuracy.

Inventive Principle:
Principle #3Local quality

2Manufacturing precision

If point cloud data is transmitted at high quality, then reconstruction fidelity is improved, but distribution efficiency deteriorates

Engineering Contradiction:
Improvereconstruction fidelityVSAvoiddistribution efficiency
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The two-layer structure enables a decoder to access only the base layer for efficient distribution scenarios, while optionally adding the enhancement layer for high fidelity requirements. This segmentation makes the system adaptable to different distribution efficiency and quality requirements simultaneously.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The enhancement layer contains more detail information than strictly necessary for basic reconstruction. By providing this excessive information selectively, the system allows receivers to achieve high fidelity when needed while maintaining efficient distribution by keeping the base layer alone for general purposes.

Inventive Principle:
Principle #16Partial or excessive action

3Manufacturing precision

If lossless compression is applied to all point cloud data, then reconstruction quality is maintained, but bit-rate increases

Engineering Contradiction:
Improvereconstruction qualityVSAvoidbit-rate
Core Design Contradiction:
Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The patent applies lossless compression only to the enhancement layer containing critical detail information, while the base layer uses lossy compression. This segmentation ensures that only the necessary bits are encoded losslessly, maintaining reconstruction quality while minimizing overall bit-rate consumption.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Lossless encoding is applied locally only to the enhancement layer where quality preservation is critical, rather than to the entire point cloud data. This local application of lossless compression maintains reconstruction quality where needed while significantly reducing the total bit-rate compared to full lossless encoding.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3922037B1Processing a point cloud
Publication Date: 2026.04.08 INTERDIGITAL VC HOLDINGS INC
  • EP3922037B1 patent drawingFigure 1~2
  • EP3922037B1 patent drawingFigure 3
  • EP3922037B1 patent drawingFigure 3a

AI summary

At least one embodiment relates to a method for transmitting PLRM metadata that removes the dependency between PLRM metadata and the occupancy map and the bloc to patch information BlockToPatch.