Point Cloud Encoding via Patch Projection and Octree

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing point cloud compression methods are inefficient for both sparse and dense point clouds, with octree-based representations being ineffective for dense clouds and segmentation-based methods introducing approximation errors and high computation complexity.

Innovation Solution

A method that segments the point cloud into plane-like patches, determines inverse projection parameters, projects these patches to planes, encodes the planes, and uses octree-decomposition to encode sparse points not belonging to any patch, combining the advantages of segmentation and octree compression techniques.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If octree-based representation is used for point cloud compression, then sparse point clouds are efficiently compressed, but dense point clouds are inefficiently represented

Engineering Contradiction:

Improvecompression efficiencyVSAvoidapplicability to dense point clouds

Core Design Contradiction:

ProductivityVSAdaptability or versatility

Solution Approach 1:

The point cloud is segmented into multiple patches based on spatial distribution characteristics. Each patch is then processed using the octree-based representation method, allowing the system to handle both sparse and dense regions effectively. This segmentation enables the compression algorithm to adapt to local density variations while maintaining overall efficiency.

Inventive Principle:

Principle #1Segmentation

2Productivity

If segmentation-based point cloud representation is used, then dense point clouds are efficiently compressed, but computation complexity increases

Engineering Contradiction:

Improvecompression efficiencyVSAvoidcomputation complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

Different compression strategies are applied to different regions of the point cloud based on their local characteristics. Sparse regions use octree-based representation while dense regions use segmentation-based methods. This local quality approach optimizes compression efficiency for each region while avoiding the need to apply complex segmentation to the entire point cloud, thereby reducing overall computation complexity.

Inventive Principle:

Principle #3Local quality

3Productivity

If segmentation-based point cloud representation is used, then dense point clouds are compressed, but approximation errors are introduced

Engineering Contradiction:

Improvecompression capabilityVSAvoidgeometric accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system dynamically selects between octree-based representation and segmentation-based representation for different patches of the point cloud. By adapting the representation method to local density characteristics, the system maintains geometric accuracy in sparse regions using octree while using segmentation only where appropriate for dense regions, thereby minimizing overall approximation errors.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS11348260B2Methods and devices for encoding and reconstructing a point cloud

Publication Date: 2022.05.31 INTERDIGITAL VC HOLDINGS INC

AI summary

This method for encoding a 3D point cloud comprises: —segmenting the point cloud into first patches (2); —determining inverse projection parameters of the patches (4); —projecting the first patches to planes using the inverse projection parameters (10); —encoding the planes (12); —determining sparse points in the point cloud not belonging to any patch (HC, 18); and—encoding the sparse points using an octree-decomposition (20).