AI-Based 2D-to-3D Mesh Generation for Dynamic Motion

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing 3D imaging technologies are insufficient in generating dynamic motion images and accurately restoring hidden or cropped object parts, and lack effective methods for calculating motion similarity and generating images from sentences.

Innovation Solution

A method and apparatus using artificial intelligence to generate three-dimensional motion images from two-dimensional images, restore hidden object parts using UV maps, and calculate motion similarity using three-dimensional mesh models, while generating images from sentences by analyzing text.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If 3D imaging technology is used to generate motion images from 2D images, then dynamic motion capability is improved, but technology completeness is insufficient

Engineering Contradiction:
Improvedynamic motion capabilityVSAvoidtechnology completeness
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The system segments the 3D imaging process into distinct modules: 2D image input, AI-based 3D mesh generation, motion detection algorithms, and rendering engine. This modular segmentation allows each component to be optimized independently while working together to achieve complete motion image generation capability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies dimensionality change by converting 2D images into 3D mesh models, adding the third dimension (depth) to static images. This enables the generation of three-dimensional motion images from two-dimensional input, fundamentally enhancing the adaptability of the imaging system.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If AI models are used to restore hidden or cropped object parts, then restoration accuracy is improved, but computational complexity increases

Engineering Contradiction:
Improverestoration accuracyVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preliminary action by pre-training AI models on extensive datasets of complete objects and their partial views. This pre-training enables the models to quickly infer and restore hidden or cropped parts with high accuracy without requiring complex real-time computations during the actual restoration process.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If three-dimensional mesh models are used to calculate motion similarity, then calculation precision is improved, but processing time increases

Engineering Contradiction:
Improvemotion similarity calculation precisionVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent extracts key motion features from three-dimensional mesh models, such as pose parameters, motion vectors, and spatial relationships. By extracting only the essential motion characteristics rather than processing the entire mesh data, the system achieves precise motion similarity calculation while significantly reducing processing time.

Inventive Principle:
Principle #2Taking out (Extraction)

4Extent of automation

If automatic motion image generation from sentences is implemented, then automation level is improved, but system complexity increases

Engineering Contradiction:
Improveautomatic image generation capabilityVSAvoidsystem complexity
Core Design Contradiction:
Extent of automationVSDevice complexity

Solution Approach 1:

The system introduces an intermediary natural language processing layer that translates human sentences into structured motion descriptors and scene parameters. This intermediary layer acts as a bridge between natural language input and the 3D motion generation engine, enabling automatic image generation from sentences while managing system complexity through standardized intermediate representations.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP4641508A1Method and system for generating three-dimensional motion object based on artificial intelligence
Publication Date: 2025.10.29 NATIONA INC
  • EP4641508A1 patent drawingFigure 1~2
  • EP4641508A1 patent drawingFigure 3
  • EP4641508A1 patent drawingFigure 4~5

AI summary

Provided are a method and apparatus for generating a three-dimensional motion image based on artificial intelligence. The method for generating a three-dimensional motion image based on artificial intelligence, according to an embodiment of the present invention, is performed by a computing apparatus and may comprise the steps of: acquiring a two-dimensional image; by using a machine-trained artificial intelligence module, identifying an object from the two-dimensional image and analyzing the motion of the identified object to generate three-dimensional modeling data of the object; and generating a three-dimensional motion image by using the three-dimensional modeling data. According to a three-dimensional motion image generation method according to the present invention, on the basis of artificial intelligence, the motion of an object may be extracted from a two-dimensional image and implemented as a three-dimensional motion image. In addition, three-dimensional mesh data of an object may be generated by using a two-dimensional image of the object.