Low-Parallax Multi-Camera Imaging System for Seamless Panoramic Compositing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Panoramic multi-camera systems face challenges in efficiently stitching and blending images due to significant parallax differences between adjacent cameras, leading to image artifacts and increased processing time, especially when capturing 360-degree panoramic images.

Innovation Solution

The use of low-parallax cameras designed to share a common center of perspective, with optimized lens systems and opto-mechanical designs to minimize parallax errors, allowing for real-time image compositing and improved image quality by reducing residual parallax to imperceptible levels.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If cameras are sparsely populating the outer surface to capture complete 360-degree panoramic images, then the field of view coverage is improved, but the parallax differences and image processing complexity increase significantly

Engineering Contradiction:

Improvefield of view coverageVSAvoidimage processing complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The camera system is divided into multiple discrete camera units positioned at specific locations around the sphere. Each camera captures a specific sector, and the segments are stitched together to form the complete panoramic view. This segmentation allows complete 360-degree coverage while managing processing complexity through structured organization.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Multiple camera images are merged and stitched together to form a complete panoramic view. The system combines images from sparsely distributed cameras by aligning and compositing them, filling gaps through overlapping fields of view to create a seamless 360-degree panorama.

Inventive Principle:

Principle #5Merging (Combining)

2Adaptability or versatility

If cameras are positioned close together to reduce gaps, then the image coverage is improved, but the parallax differences between adjacent cameras increase

Engineering Contradiction:

Improveimage coverageVSAvoidparallax difference

Core Design Contradiction:

Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The system uses different camera configurations and field of view parameters for different regions. Adjacent cameras have their FOVs specifically adjusted to overlap by controlled amounts (5-20%), and the parallax compensation is locally optimized for each camera position to minimize discrepancies in the final stitched image.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system dynamically adjusts camera parameters including field of view angle, exposure settings, and focus distances for each camera based on its specific position. This parameter optimization reduces parallax differences by ensuring consistent imaging characteristics across adjacent cameras while maintaining close spacing for comprehensive coverage.

Inventive Principle:

Principle #35Parameter changes

3Adaptability or versatility

If cameras have widened fields of view to capture gaps between adjacent cameras, then the image coverage is improved, but the parallax differences and processing time increase

Engineering Contradiction:

Improveimage coverageVSAvoidprocessing time

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

Cameras are positioned with slight overlaps in their fields of view, capturing slightly more area than strictly necessary to fill gaps. This partial excessive action ensures complete coverage without requiring excessive FOV widening, thereby limiting the increase in processing time while maintaining image quality.

Inventive Principle:

Principle #16Partial or excessive action

4Manufacturing precision

If adjacent low parallax cameras are assembled together to produce real-time composite panoramic images, then the image quality is improved, but the calibration and assembly processes become more complex

Engineering Contradiction:

Improveimage qualityVSAvoidcalibration process complexity

Core Design Contradiction:

Manufacturing precisionVSDevice complexity

Solution Approach 1:

Each camera is pre-calibrated individually with its intrinsic parameters (focal length, optical center, distortion coefficients) determined before assembly. The relative positions and orientations of cameras are also pre-determined through extrinsic calibration. This preliminary action simplifies the final composite image generation by having all calibration data ready in advance.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses automatic calibration methods where cameras capture images of known patterns or use feature detection algorithms to self-calibrate their positions and orientations. This self-service approach reduces manual calibration complexity while ensuring accurate geometric relationships between cameras for high-quality composite image production.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS20250022103A1Image Compositing with Adjacent Low Parallax Cameras

Publication Date: 2025.01.16 CIRCLE OPTICS INC

AI summary

A low-parallax multi-camera imaging system may enable combination of images from multiple camera channels into a panoramic image. In some examples, the imaging system may be designed to include small areas of overlap between adjacent camera channels. Panoramic images may be generated by compositing image data from multiple camera channels by using techniques described herein. In some examples, contribution of each camera channel may be weighted based on factors such as distances relative to an overlap region or content within the overlap region.