Daisy-Chained Thermal Cameras for RGB-Thermal Distortion Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Thermal camera systems face a tradeoff between field-of-view (FoV) and image quality, requiring multiple cameras and numerous wires, which are costly and prone to tangling, while existing methods fail to efficiently map RGB images to thermal images in low-quality, wide FoV devices.
Innovation Solution
A method and system for generating a one-to-one mapping of fiducials in RGB and thermal images, accounting for distortions, using a daisy chain of low-quality, wide FoV thermal cameras with polynomial and affine transformations, reducing the need for high-quality cameras and wires.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple high-quality thermal cameras are used to achieve high image quality, then image quality is improved, but device complexity and cost increase
Solution Approach 1:
The patent divides the imaging task into two separate cameras: one dedicated to capturing RGB images and another dedicated to capturing thermal images. This segmentation allows each camera to be optimized for its specific function, using lower-cost, lower-quality sensors while achieving the overall system goal of capturing both visual and thermal data effectively
Solution Approach 2:
The patent creates a mapping relationship between the RGB image space and thermal image space by identifying corresponding fiducial markers in both images. This copying approach allows the system to transfer spatial information from the high-quality RGB camera to the thermal camera data, effectively enhancing the thermal image quality without requiring a high-quality thermal camera
2Area of stationary object
If multiple thermal cameras are linked together to achieve wide field-of-view, then field-of-view is improved, but device complexity and wire management increase
Solution Approach 1:
The patent segments the imaging functions by using one camera for RGB capture and another for thermal capture, each with their own field-of-view characteristics. This allows the system to achieve comprehensive coverage through functional division rather than requiring multiple cameras of the same type linked together with extensive wiring
Solution Approach 2:
The patent makes each camera multi-functional by having the RGB camera capture both visual information and spatial reference data (through fiducial markers), and the thermal camera capture both thermal information and corresponding spatial data. This universality reduces the need for additional cameras and wiring while achieving comprehensive environmental monitoring
3Device complexity
If low-quality wide FoV thermal cameras are used, then device complexity is reduced, but image distortion increases
Solution Approach 1:
The patent copies the spatial geometry information from the undistorted RGB image to the thermal image by establishing correspondence through fiducial markers. This allows the thermal image to inherit the accurate spatial relationships from the RGB image, compensating for the inherent distortions in the low-quality wide-FoV thermal camera
Solution Approach 2:
The patent applies parameter transformations through mapping algorithms that adjust the thermal image data based on the geometric relationship established by fiducial markers. This transforms the distorted thermal image coordinates into accurate spatial coordinates, correcting distortion without requiring hardware changes
Data Source
AI summary
A thermal camera system configured for Red-Green-Blue (RGB) to thermal image mapping and calibration of a thermal camera. The system may comprise a plurality of thermal cameras connected in a daisy-chain formation, and a computing device communicatively coupled to the base thermal camera. The computing device configured to accept a distorted RGB image, convert it into an array image, undistort the array image into an undistorted RGB image through use of a barrel transformation, map each corner element of the plurality of corner elements to a predefined coordinate to generate a thermal angular mapping, and map the thermal angular mapping to a distorted thermal image by a 2nd-degree parabola mapping process.