AR Rendering from 2D Drawings with Real-Time 3D Model Generation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing augmented reality systems fail to accurately generate three-dimensional models from two-dimensional illustrations in real-time.

Innovation Solution

A system that performs camera calibration, calculates camera pose, detects and processes 2D drawings to generate a 3D model, and renders an augmented reality environment in real-time, using computer vision techniques and libraries like OpenCV.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If existing augmented reality systems process 2D images, then image processing capability is provided, but accurate 3D model generation from 2D illustrations in real-time is not achieved

Engineering Contradiction:
Improve3D model accuracyVSAvoidreal-time processing speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The system segments the complex task of 3D model generation into distinct processing stages: camera calibration, camera pose calculation, 2D drawing detection and processing, 3D model generation, and rendering. Each stage is handled by specialized modules that can operate independently and in parallel where possible, enabling real-time processing while maintaining high accuracy in each individual stage.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs camera calibration as a preliminary action before processing 2D illustrations. By pre-determining camera parameters and transformation matrices, the system eliminates the need for complex real-time calibration calculations, thereby improving both the accuracy of 3D model generation and the speed of real-time processing.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If camera calibration and pose calculation are performed, then accurate object positioning is achieved, but processing time increases

Engineering Contradiction:
Improveobject position accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

Camera calibration is performed as a preliminary action that establishes transformation matrices and camera parameters in advance. These pre-calculated values are then reused during real-time processing, eliminating the need for repeated calibration calculations and significantly reducing processing time while maintaining high positioning accuracy.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system creates a mathematical model (copy) of the camera's transformation behavior through calibration. This model, represented by transformation matrices and intrinsic parameters, allows the system to accurately determine object positions without performing complex geometric calculations for each new image, thus reducing processing time while maintaining precision.

Inventive Principle:
Principle #26Copying

3Loss of information

If complete 3D model generation is implemented, then model completeness is improved, but computational complexity increases

Engineering Contradiction:
Improvemodel completenessVSAvoidsystem complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The 3D model generation process is segmented into distinct computational stages: 2D feature detection, 3D point cloud generation, mesh construction, and rendering. Each stage processes only the necessary data for that specific purpose, avoiding redundant calculations and reducing overall system complexity while ensuring complete 3D model generation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system efficiently handles the transition from 2D to 3D by working in the appropriate dimensionality at each stage. 2D image processing operations are performed in two dimensions, then results are transformed into 3D space only when necessary for model construction. This dimensional approach reduces computational complexity by avoiding premature 3D calculations while still achieving complete 3D model generation.

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

Data Source

PatentEP3881295B1Systems and methods for generating augmented reality environments from two-dimensional drawings
Publication Date: 2026.04.08 GEOMNI INC
  • EP3881295B1 patent drawingFigure 1
  • EP3881295B1 patent drawingFigure 2
  • EP3881295B1 patent drawingFigure 3

AI summary

Systems and methods for generating augmented reality environments from 2D drawings are provided. The system performs a camera calibration process to determine how a camera transforms images from the real world into a 2D image plane. The system calculates a camera pose and determines an object position and an object orientation relative to a known coordinate system. The system detects and processes a 2D drawing/illustration and generates a 3D model from the 2D drawing/illustration. The system performs a rendering process, wherein the system generates an augmented reality environment which includes the 3D model superimposed on an image of the 2D drawing/illustration. The system can generate the augmented reality environment in real time, allowing the system to provide immediate feedback to the user. The images processed by the system can be from a video, from multiple image photography, etc.