3D Model Compression via Instance Transformation Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current 3D model compression methods fail to effectively handle 3D model properties like normal, color, and texture coordinates, especially when instance transformations include reflection, leading to inefficient compression and decoding processes.

Innovation Solution

A method and apparatus for generating a compressed bitstream that separates and compresses instance transformations into reflection, rotation, translation, and scaling parts, with options for elementary and grouped instance data modes, and uses Cartesian or spherical representations for rotation, allowing for flexible compression and decoding strategies.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If current 3D model compression methods are used, then compression is achieved, but they fail to handle 3D model properties like normal, color, and texture coordinates when instance transformations include reflection

Engineering Contradiction:
Improvehandling of 3D model propertiesVSAvoidcompression effectiveness
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The instance transformation is segmented into four distinct parts: reflection part, rotation part, translation part, and scaling part. Each part is compressed separately with dedicated data fields, allowing the system to handle diverse 3D model properties (normal, color, texture coordinates) and transformation types effectively without compromise in compression effectiveness.

Inventive Principle:
Principle #1Segmentation

2Ease of manufacture

If instance transformations are compressed as a whole, then compression is simpler, but reflection transformations cannot be handled properly

Engineering Contradiction:
Improvecompression simplicityVSAvoidreflection transformation handling
Core Design Contradiction:
Ease of manufactureVSAdaptability or versatility

Solution Approach 1:

The transformation compression is divided into separate data fields for reflection, rotation, translation, and scaling. This segmentation maintains relative simplicity through structured organization while simultaneously enabling proper handling of reflection transformations and other transformation types that would be impossible to handle with a monolithic approach.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The compression system dynamically adapts to different transformation types by providing separate data fields that can be selectively applied. This allows the system to handle reflection, rotation, translation, and scaling transformations appropriately based on the specific instance requirements, achieving both simplicity and versatility.

Inventive Principle:
Principle #15Dynamics

3Measurement precision

If more detailed transformation data is stored, then decoding accuracy improves, but bitstream size increases

Engineering Contradiction:
Improvedecoding accuracyVSAvoidbitstream size
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

By segmenting transformation data into distinct parts (reflection, rotation, translation, scaling), the system can apply appropriate precision to each component based on its requirements. This allows maintaining high decoding accuracy for critical components while optimizing bitstream size by not over-specifying precision for less critical components.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different data fields are used for different transformation components, allowing local optimization of data precision. Critical transformation parameters can be stored with higher precision while less critical parameters use compact representations, achieving the balance between decoding accuracy and bitstream size.

Inventive Principle:
Principle #3Local quality

4Adaptability or versatility

If separate compression modes are provided, then adaptability to different applications improves, but system complexity increases

Engineering Contradiction:
Improveapplication adaptabilityVSAvoidcompression system complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The compression system provides two operational modes (elementary instance data mode and grouped instance data mode) that can be dynamically selected based on application requirements. This dynamic flexibility achieves high adaptability while managing complexity through a unified underlying data structure that supports both modes.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The same set of transformation data fields (reflection, rotation, translation, scaling) serves multiple purposes and supports both compression modes universally. This multi-functionality allows the system to adapt to different applications without requiring separate complex systems for each mode, thereby managing complexity while maintaining versatility.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP2783509B1Method and apparatus for generating a bitstream of repetitive structure discovery based 3D model compression
Publication Date: 2024.10.30 INTERDIGITAL VC HOLDINGS INC
  • EP2783509B1 patent drawingFigure 1~2
  • EP2783509B1 patent drawingFigure 3~4
  • EP2783509B1 patent drawing

AI summary

A method and apparatus for generating a bitstream representative of a 3D model, and a method and an apparatus for processing the same. A 3D model is modeled by using a using a 'pattern-instance' representation, wherein a pattern is a representative geometry of a repetitive structure, and the connected components belonging to the repetitive structure is call an instance of the corresponding pattern. After discovery of the repetitive structures and their transformations and properties, the present embodiments provide for generating a bitstream in either a first format or a second format. In the first format, the pattern ID and its associated transformation and property information are grouped together in the bitstream, and in the second format the pattern ID, transformation property and property information are grouped together according to information type.