Panoramic 3D Camera Array for Parallax Error Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional 3D cameras have limited field-of-view and suffer from parallax errors when attempting to create panoramic 3D images or videos, leading to degraded image quality due to the need for iterative processing and significant distortions with ultra-wide-angle lenses.
Innovation Solution
A system of multiple stereoscopic pairs of compound/composite cameras arranged in various geometries to capture overlapping images, which are then stitched to generate panoramic 3D images or videos with a 360° field of view, reducing parallax errors through techniques described in co-pending patent applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of moving object
If conventional stereoscopic cameras are used to capture 3D images, then the field of view is limited to approximately 85°, but the image quality is maintained
Solution Approach 1:
The imaging system is divided into multiple camera sensors (at least three) arranged in an array along a convex axis, with each sensor capturing a portion of the panoramic scene. This segmentation allows the system to achieve a wide field of view while maintaining image quality through coordinated capture and stitching of multiple views.
2Area of moving object
If ultra-wide-angle lenses are used to capture panoramic videos, then the field of view is expanded, but significant radial and scale distortions occur
Solution Approach 1:
Instead of using a single ultra-wide-angle lens that causes distortion, the system segments the panoramic capture into multiple overlapping views from regular FOV cameras arranged in an array. This avoids the distortion problem while achieving the same panoramic effect through geometric arrangement and stitching.
Solution Approach 2:
The system transitions from a single-lens approach to a multi-sensor array arrangement in three-dimensional space along a convex axis. This spatial dimensionality change allows panoramic capture without relying on wide-angle lens distortion, using instead the geometric relationships between multiple sensors.
3Area of moving object
If multiple 3D camera pairs are used to generate panoramic 3D video, then the field of view is expanded, but parallax error increases significantly with distance between camera pairs
Solution Approach 1:
The panoramic 3D capture is segmented into multiple stereoscopic pairs of compound cameras, with each pair capturing a localized 3D view. By distributing these pairs along a convex axis and using overlapping fields of view, the system achieves panoramic coverage while minimizing parallax errors through the stitching process.
Solution Approach 2:
The system arranges multiple stereoscopic camera pairs in a three-dimensional geometric configuration along a convex axis rather than spreading them linearly over large distances. This spatial arrangement reduces the baseline distance between corresponding sensors in different pairs, thereby minimizing parallax errors while still achieving panoramic coverage.
4Area of moving object
If iterative processing is used to create panoramic 3D images from multiple camera pairs, then panoramic coverage is achieved, but image quality is measurably degraded
Solution Approach 1:
The system performs preliminary synchronization of multiple camera sensors to capture images simultaneously, establishing precise geometric relationships before stitching. This preliminary coordination reduces the need for complex iterative processing and maintains image quality by minimizing alignment errors.
Solution Approach 2:
The system replaces complex iterative image processing with a more direct stitching approach based on the known geometric arrangement of sensors along a convex axis. By using the predetermined spatial relationships between sensors, the system reduces computational iterations and preserves image quality.
Data Source
AI summary
A three-dimensional imaging system for generating a 3D panoramic video/image is disclosed. The three-dimensional imaging system includes a plurality of stereoscopic pairs of compound cameras which are separated by a predefined distance d. The imaging system is configured to operate in a first mode to generate panoramic images simultaneously using the compound cameras. Spatial distortion and temporal distortion are calculated for each column in a region of overlap of the field of view of the panoramic images generated. The column for which the spatial and temporal distortion is minimum is marked as a seam location. The marked seam location is changed based on location of a motion in the column for which the calculated spatial and temporal distortion is minimum.