Thermal Pixel Classification Using Background Intensity Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing thermal imaging techniques struggle to consistently determine a reliable digitisation image threshold value due to insufficient measurement data points, particularly in arrays with a low number of pixels, leading to inaccurate differentiation between background and thermal objects.
Innovation Solution
A thermal imaging apparatus and method utilizing an array of thermal sensing pixels with signal processing circuitry that includes a background identifier and pixel classifier, which identifies a largest number of substantially the same pixel intensity values to generate a model of expected background intensity levels, and classifies pixel measurements as background or object based on temperature threshold values adjusted by a noise margin.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If histogram-based techniques with data binning are used to determine the digitisation threshold value, then the method can be applied with limited measurement data points, but the results are not consistently reliable
Solution Approach 1:
The patent introduces an intermediary statistical model that mediates between the limited raw measurement data and the threshold determination process. This model incorporates prior knowledge about the expected distributions of background and foreground pixel intensities, acting as a bridge that stabilizes threshold calculation even when measurement data is scarce. The model effectively smooths over the insufficiency of data points by leveraging the statistical intermediary framework.
Solution Approach 2:
The patent performs preliminary action by pre-establishing statistical models and expected distribution characteristics before actual threshold determination. The system pre-calculates or pre-defines the expected intensity distributions for background and foreground regions, so that when limited measurement data is available, these pre-established models can guide the threshold selection process, ensuring consistency even with insufficient data points.
2Device complexity
If a low pixel array (e.g., 24 x 32 pixels) is used in the thermal detector device, then the device complexity and cost are reduced, but insufficient measurement data points are available for accurate threshold determination
Solution Approach 1:
The patent changes the parameter space by moving from direct pixel intensity analysis to a statistical model-based approach. Instead of relying on the raw number of pixels, the system transforms the problem into estimating threshold values based on statistical distributions and expected intensity patterns. This parameter transformation allows accurate threshold determination even when the physical pixel array size is small, effectively decoupling measurement precision from device complexity.
Solution Approach 2:
The patent creates a statistical copy or representation of the expected background and foreground intensity distributions. Rather than requiring numerous physical pixels to directly measure all variations, the system creates statistical models that copy the essential characteristics of what the intensity distributions should look like. This statistical copying allows the system to infer threshold information that would otherwise require many more physical measurement points.
3Device complexity
If traditional histogram techniques are used with few data bins, then processing is simpler, but the digitisation threshold value cannot be reliably determined
Solution Approach 1:
The patent adds another dimension to the analysis by incorporating statistical distribution modeling and expected intensity patterns alongside the traditional histogram approach. Instead of relying solely on the single dimension of binned frequency counts, the system adds the dimension of statistical expectation and distribution fitting. This multi-dimensional approach maintains processing simplicity while significantly improving threshold determination reliability through the combined use of empirical data and statistical models.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Improves the accuracy and consistency of background detection in thermal imaging, even with low pixel arrays, by effectively distinguishing between background and objects, enhancing the reliability of thermal image processing.
Implementation Method 1
a thermal detector device comprising an array of thermal sensing pixels that respectively receive and measure infra-red radiation
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
A thermal imaging apparatus comprising: a thermal detector device (100) comprising an array of thermal sensing pixels (102) and signal processing circuitry (104) coupled to the detector device (100). The circuitry (104) supports a background identifier (110) and a pixel classifier (112), the background identifier (110) comprising a common intensity identifier (114) and an expected background intensity calculator (116). The background identifier (110) receives pixel measurement data captured by the detector device (100) in respect of pixels of the array (102) and the common intensity identifier (114) identifies a largest number of substantially the same pixel intensity values from the pixel measurement data. The expected background intensity calculator (116) uses the largest number of substantially the same pixel intensity values to generate a model of expected background intensity levels. The pixel classifier (112) uses the model to determine whether an intensity measurement by a pixel (118) of the array (102) corresponds to a background or an object in an image.