360-Degree 3D Image Generation Using Fisheye Camera Distortion Correction
Find Innovative SolutionsGenerate 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
Engineering 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
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.
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.
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
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.
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.
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
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.
Data Source
Figure 1
Figure 2
Figure 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.