3D Point Cloud Depth Projection Encoding for Complex Geometry Compression

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image-based point cloud compression techniques struggle with low compression efficiency and quality when projecting point clouds with complex geometries, such as those with many folds or non-surface distributions, leading to increased bitrates and inaccurate reconstructions.

Innovation Solution

A method that determines whether to store minimum or maximum depth values in the first depth image based on the characteristics of each connected component, optimizing projection modes to better capture the geometry and reduce noise, thereby improving HEVC encoding performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If orthogonal projection of point cloud is used for compression, then compression efficiency is improved, but quality and accuracy of geometry reconstruction deteriorates for complex geometries

Engineering Contradiction:
Improvecompression efficiencyVSAvoidgeometry reconstruction accuracy
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The point cloud is divided into multiple connected components based on spatial proximity and geometric characteristics. Each connected component is processed independently with optimized projection parameters, allowing complex geometries to be segmented into manageable parts that can be accurately reconstructed while maintaining overall compression efficiency.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different projection modes and parameter settings are applied to different connected components based on their local geometric characteristics. Regions with complex folds or non-surface distributions use specialized handling, while simpler regions use standard projection, optimizing both compression efficiency and local reconstruction accuracy.

Inventive Principle:
Principle #3Local quality

2Device complexity

If standard depth image encoding is used, then encoding complexity is reduced, but compression efficiency and visual quality deteriorate

Engineering Contradiction:
Improveencoding complexityVSAvoidcompression efficiency
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The encoding process dynamically adapts to the characteristics of each connected component by selecting appropriate projection modes and parameter combinations. This dynamic approach allows the system to optimize compression efficiency for each region without requiring overly complex global processing, balancing encoding complexity with compression performance.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

Different encoding parameters such as projection direction, depth quantization levels, and transformation matrices are adjusted based on the geometric properties of each connected component. These parameter changes enable better compression efficiency and visual quality while keeping the base encoding framework relatively simple.

Inventive Principle:
Principle #35Parameter changes

3Device complexity

If simple projection schemes are used, then encoding complexity is reduced, but quality of experience and reconstruction accuracy deteriorate

Engineering Contradiction:
Improveencoding complexityVSAvoidquality of experience
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

By segmenting the point cloud into connected components, the system can apply enhanced projection techniques only where needed rather than across the entire cloud. This maintains relatively simple overall encoding while improving reconstruction accuracy for complex regions that most impact quality of experience.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system applies higher-quality projection and reconstruction techniques locally to connected components with complex geometries, while using simpler methods for regular regions. This localized enhancement improves overall quality of experience without requiring complex processing everywhere, maintaining a balance between complexity and quality.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3777183B1A method for encoding depth values of a set of 3D points once orthogonally projected into at least one image region of a projection plane
Publication Date: 2026.01.28 INTERDIGITAL VC HOLDINGS INC
  • EP3777183B1 patent drawingFigure 1
  • EP3777183B1 patent drawingFigure 2
  • EP3777183B1 patent drawingFigure 3

AI summary

At least one embodiment relates to a method comprising encoding a projection mode associated with an image region of a first depth image, said projection mode indicating if said image region stores either the minimum or the maximum depth value of at least one orthogonally projected 3D point of the point cloud along a same projection direction; and encoding said depth minimum and maximum depth values as a function of said projection mode.