3D Model Synthesis Using Database Matching From Few Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AI models for generating 3D models from 2D images require a large number of images, which is inconvenient for users and inefficient.
Innovation Solution
A method and electronic device that utilize a 3D model database to generate 3D models by determining similar models, performing deformations and mergings, and iteratively supplementing information to achieve high similarity with a small number of images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If an AI model generates a 3D model using dozens of 2D pictures, then the accuracy and completeness of the 3D model is improved, but the complexity of operation and time consumption increases significantly
Solution Approach 1:
The system pre-generates multiple 3D models from different viewpoints using AI technology before the user actually needs them. These pre-generated models are stored in a database, so when a user requests a 3D model, the system can quickly retrieve and synthesize from these pre-prepared models rather than generating from scratch using dozens of images in real-time.
Solution Approach 2:
The system creates multiple copies of 3D models from different viewpoints and stores them in a database. When generating a final 3D model, the system synthesizes by combining these pre-existing model copies rather than creating new geometry from dozens of 2D images, significantly reducing the operational burden on users.
2Manufacturing precision
If an AI model generates a 3D model using dozens of 2D pictures, then the quality of the 3D model is improved, but the time required for generation increases
Solution Approach 1:
The system performs the computationally intensive AI model generation in advance, creating multiple 3D models from different viewpoints and storing them in a database. This preliminary action shifts the time consumption from real-time generation to offline pre-processing, making the actual user-facing generation process much faster.
Solution Approach 2:
By creating and storing multiple copies of 3D models from different viewpoints in advance, the system avoids the need to perform time-consuming image processing and 3D reconstruction from dozens of images when a user requests a model. The synthesis process combines these pre-existing copies efficiently.
3Manufacturing precision
If a 3D model database stores multiple 3D models for synthesis, then the accuracy of the generated model is improved, but the device complexity increases
Solution Approach 1:
The 3D model database serves multiple functions: it stores pre-generated AI models, provides reference models for synthesis, and enables quick retrieval of models from different viewpoints. This multi-functional database reduces the need for separate systems for model generation, storage, and retrieval, thereby managing complexity while improving accuracy.
Solution Approach 2:
The system stores multiple copies of 3D models representing the same object from different viewpoints in the database. These copies are synthesized to create the final model, improving accuracy without requiring a complex real-time generation system, as the heavy lifting was done during pre-processing.
Data Source
AI summary
A method of generating a three-dimensional (3D) model and a device for generating a 3D model are provided. The method includes generating a base 3D model of an object, determining a first target 3D model similar to the base 3D model from a 3D model database storing a plurality of 3D models, generating a first synthesized 3D model, determining a first similarity between the first synthesized 3D model and the at least one object image, determining a component having a low similarity between the first synthesized 3D model and the at least one object image, determining, as a second target 3D model, a 3D model having a high similarity to the first synthesized 3D model with respect to the component having the low similarity, from the 3D model database, and generating a second synthesized 3D model.