360-Degree 3D Image Generation Using Fisheye Camera Distortion Correction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing technologies for generating 360-degree 3D stereoscopic images using multiple cameras, such as fisheye lenses, suffer from image distortion, resulting in a low stereoscopic effect, particularly in regions with limited overlapping images.

Innovation Solution

An electronic device and image processing method that extracts distortion-free sub-images from multiple cameras, maps them to spherical images, and extracts depth information to generate a high-quality, distortion-free 360-degree 3D stereoscopic image by using a processor to divide and process images from fisheye cameras with overlapping fields of view.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Area of moving object

If multiple fisheye cameras are used to generate 360-degree 3D stereoscopic images, then the field of view coverage is improved, but image distortion occurs in the depth reconstruction stage

Engineering Contradiction:
Improvefield of view coverageVSAvoidimage distortion
Core Design Contradiction:
Area of moving objectVSManufacturing precision

Solution Approach 1:

The patent divides the 360-degree spherical image into multiple distortion-free sub-images, each covering a specific field of view region. By segmenting the overall 360-degree view into smaller manageable portions, the system can process each region separately to minimize distortion while maintaining comprehensive coverage.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different processing methods to different regions of the image. Specifically, it identifies low distortion regions and pin cushion regions, and applies appropriate depth extraction and image mapping techniques to each region type to optimize the stereoscopic effect locally.

Inventive Principle:
Principle #3Local quality

2Area of moving object

If fisheye lenses are used to capture 360-degree images, then the imaging range is expanded, but the stereoscopic effect is reduced due to image distortion

Engineering Contradiction:
Improveimaging rangeVSAvoidstereoscopic effect
Core Design Contradiction:
Area of moving objectVSReliability

Solution Approach 1:

The patent maps 2D fisheye images onto a 3D spherical coordinate system, then extracts depth information along the radial dimension from the spherical image. This dimensional transformation allows the system to maintain the wide imaging range of fisheye lenses while recovering depth information necessary for stereoscopic effect.

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

Solution Approach 2:

The patent introduces spherical images as an intermediary representation between the captured fisheye images and the final 3D stereoscopic output. The spherical image serves as a distortion-free intermediate format that preserves geometric relationships, enabling accurate depth extraction and stereoscopic reconstruction.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Device complexity

If a small number of cameras are used to miniaturize the system, then device complexity is reduced, but overlapping image coverage is insufficient for depth reconstruction

Engineering Contradiction:
Improvesystem miniaturizationVSAvoidoverlapping image coverage
Core Design Contradiction:
Device complexityVSLoss of information

Solution Approach 1:

The patent utilizes spherical geometry to organize and overlap the fields of view of multiple cameras. By mapping camera views onto a spherical coordinate system, the system achieves optimal overlapping coverage with fewer cameras, as the spherical arrangement maximizes the interlacing of视 fields compared to planar arrangements.

Inventive Principle:
Principle #14Spheroidality (Curvature)

Data Source

PatentEP3334156B1Electronic device for generating 360 degree three-dimensional image, and method therefor
Publication Date: 2021.03.31 SAMSUNG ELECTRONICS CO LTD
  • EP3334156B1 patent drawingFigure 1
  • EP3334156B1 patent drawingFigure 2
  • EP3334156B1 patent drawingFigure 3

AI summary

The present disclosure relates to an electronic device and an image processing method thereof which image a plurality of images using a plurality of cameras, generate a left-eyed spherical image and a right-eyed spherical image by dividing the plurality of images into left-eyed images and right-eyed images, acquire depth information using the generated left- and right-eyed spherical images, and generate a 360-degree 3D stereoscopic image of which a stereoscopic effect is controlled using the acquired depth information.