Pixel Imbalance Correction in Thermal Imaging Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image recording systems, such as CCD cameras, face challenges in achieving homogeneous radiation sources for accurate temperature distribution measurements due to differences in pixel recording characteristics, leading to measurement uncertainties and difficulties in calibration, especially with high-precision thermal cameras.
Innovation Solution
A method involving a radiation source system that records an object point with three image points to calculate correction factors, allowing for the control of radiation source zones to achieve a predetermined homogeneous radiation distribution, thereby reducing measurement uncertainty and improving calibration accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a planar radiator with homogeneous radiance distribution is used for calibration, then calibration simplicity is improved, but manufacturability deteriorates
Solution Approach 1:
The patent changes the parameter of radiance distribution from homogeneous to inhomogeneous, and changes the shape parameter from planar to three-dimensional (e.g., hemispherical). This allows the radiator to be manufacturable while still enabling calibration through computational correction of the known inhomogeneous distribution pattern.
2Measurement precision
If each individual pixel is illuminated with a time-invariant radiation source to determine correction factor, then measurement precision is improved, but calibration time deteriorates
Solution Approach 1:
The patent segments the calibration process into two parts: first, capture a single image with all pixels exposed to the inhomogeneous radiator; second, computationally determine correction factors by comparing each pixel's response to the known radiance distribution at its position. This parallel processing approach maintains precision while reducing calibration time from sequential to simultaneous measurement.
Solution Approach 2:
The patent uses a known theoretical or measured radiance distribution pattern as a reference model (copy) to compare against actual pixel measurements. This allows determination of correction factors from a single image capture event, eliminating the need for time-consuming sequential pixel calibration.
3Measurement precision
If pixels are calibrated using recordings at two different illumination levels, then transmission characteristic correction is improved, but ease of operation deteriorates when using constant light sources
Solution Approach 1:
The patent changes the approach from varying illumination levels to varying spatial positions. By moving the radiator or changing the optical path, different regions of the inhomogeneous radiator illuminate different pixels, effectively providing multiple illumination conditions from a single static image capture. This maintains correction precision while simplifying operation.
Data Source
Figure 1a~2c
Figure 3
Figure 4
AI summary
The invention relates to a method for determining the imbalance of the transmission behaviour of individual or all pixels (32) of an image capturing system (12), wherein the image capturing system (12) has a plurality of pixel columns and a plurality of pixel rows formed by pixels (32), and/or for determining the radiance distribution of the radiation source using the following steps: (a) capturing a radiation source (10) with sufficiently temporally constant radiance using the image capturing system (12), such that a primary image is obtained; (b) capturing the radiation source (10) with column-offsetting, such that a column-offset image is obtained; (c) capturing the radiation source (10) with row-offsetting, such that a row-offset image is obtained; and (d) calculating a correction factor matrix from the primary image, the row-offset image and the column-offset image, containing the correction parameters for correcting measured values of the pixels (32).