3D Scene Reconstruction from 2D Images for VR

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Two-dimensional images fail to provide a full three-dimensional experience, limiting the ability to relive the sensation of a captured scene, and carrying bulky 3D cameras is impractical for everyday use.

Innovation Solution

A method to generate a three-dimensional image from a set of two-dimensional input images taken from different vantage points, using sparse and dense reconstruction representations, depth testing, and multi-layered geometric mesh rendering to create a immersive 3D scene.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If two-dimensional images are used to capture scenes, then the ease of operation is improved, but the immersion experience deteriorates

Engineering Contradiction:
Improveease of captureVSAvoidthree-dimensional experience
Core Design Contradiction:
Ease of operationVSLoss of information

Solution Approach 1:

The patent transforms two-dimensional images into a three-dimensional representation by generating depth maps and constructing a multi-layered geometric mesh. This dimensionality change allows the system to preserve spatial information and provide immersive 3D experiences while maintaining the simplicity of using standard 2D cameras for capture.

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

2Loss of information

If three-dimensional cameras are used to capture scenes, then the three-dimensional experience is improved, but the device complexity deteriorates

Engineering Contradiction:
Improvethree-dimensional experienceVSAvoidcamera equipment
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

Instead of using complex 3D cameras, the patent creates a computational copy of the 3D scene by processing multiple 2D images through sparse and dense reconstruction algorithms. This generates a multi-layered geometric mesh that replicates the three-dimensional structure, allowing standard cameras to achieve 3D capture capabilities through software processing.

Inventive Principle:
Principle #26Copying

3Manufacturing precision

If multiple processing steps are applied to generate 3D reconstruction, then the three-dimensional quality is improved, but the processing time deteriorates

Engineering Contradiction:
Improvereconstruction qualityVSAvoidprocessing time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent divides the complex 3D reconstruction process into distinct segments: sparse reconstruction to establish initial 3D points, dense reconstruction to generate detailed depth maps, and mesh generation to create the final geometric model. This segmentation allows each stage to be optimized independently and processed efficiently, balancing quality with processing time.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3349176B1Three-dimensional scene reconstruction from set of two-dimensional images for consumption in virtual reality
Publication Date: 2021.05.12 META PLATFORMS INC
  • EP3349176B1 patent drawingFigure 1
  • EP3349176B1 patent drawingFigure 2
  • EP3349176B1 patent drawingFigure 3

AI summary

To enable better sharing and preservation of immersive experiences, a graphics system reconstructs a three-dimensional scene from a set of images of the scene taken from different vantage points. The system processes each image to extract depth information therefrom and then stitches the images (both color and depth information) into a multi-layered panorama that includes at least front and back surface layers. The front and back surface layers are then merged to remove redundancies and create connections between neighboring pixels that are likely to represent the same object, while removing connections between neighboring pixels that are not. The resulting layered panorama with depth information can be rendered using a virtual reality (VR) system, a mobile device, or other computing and display platforms using standard rendering techniques, to enable three-dimensional viewing of the scene.