A method and apparatus for estimating temperature measurement errors of an infrared thermal imager

Abstract

The application discloses a temperature measurement error estimation method and device for an infrared thermal imager. The method comprises the following steps: setting a discrete temperature value set of a blackbody radiation source; based on the discrete temperature value set, performing temperature measurement on the blackbody radiation source by using the infrared thermal imager to obtain a measurement temperature set; and processing the measurement temperature set to obtain a measurement temperature correction model of the infrared thermal imager.

Description

Technical Field

[0001] This invention relates to the fields of industrial data mining and processing, infrared image measurement, and strategy optimization, specifically to a method and apparatus for estimating the temperature measurement error of an infrared thermal imager. Background Technology

[0002] An infrared thermal imager is a device that uses an infrared optical system, an infrared detector, and an electronic processing system to convert the infrared thermal radiation emitted from an object's surface into a visible light image and display it on a screen. In recent years, infrared thermal imagers have been widely used in fields such as security monitoring and medical detection. Therefore, the correction of temperature measurement errors in infrared thermal imagers is of great significance to their development.

[0003] Existing error correction methods for infrared thermal imagers mostly rely on measurement data from a single temperature point or a limited number of pixels to construct correction models, neglecting the response differences between different pixels and the uniformity of temperature distribution across the entire field of view. Some methods employ a single regression algorithm for modeling, which struggles to adapt to complex error sources, resulting in insufficient robustness and generalization ability of the correction model. Furthermore, outliers in the measurement data are not effectively removed, further impacting the accuracy of the correction model. In addition, the fusion methods of traditional correction models are relatively simple, failing to fully leverage the advantages of different regression algorithms and thus unable to achieve precise error compensation. These problems make existing correction methods insufficient for high-precision temperature measurement scenarios, limiting the widespread adoption of infrared thermal imagers in high-end applications. Therefore, a technical solution that can comprehensively improve the accuracy of temperature measurement error estimation and the reliability of correction is urgently needed. Summary of the Invention

[0004] This invention primarily addresses the problem of how to accurately estimate the temperature measurement error of an infrared thermal imager and comprehensively improve the accuracy of temperature measurement error estimation and the reliability of correction. This invention discloses a method and apparatus for estimating the temperature measurement error of an infrared thermal imager.

[0005] In a first aspect, this invention discloses a method for estimating the temperature measurement error of an infrared thermal imager, comprising:

[0006] S1 sets the discrete temperature value set of the blackbody radiation source;

[0007] S2, Based on the set of discrete temperature values, use an infrared thermal imager to measure the temperature of the blackbody radiation source to obtain a set of measured temperatures;

[0008] S3, process the measured temperature set to obtain the measurement temperature correction model of the infrared thermal imager.

[0009] The method of measuring the temperature of a blackbody radiation source using an infrared thermal imager based on the discrete temperature value set to obtain a set of measured temperatures includes:

[0010] S21, Set up a blackbody radiation source, which is placed under each discrete temperature value in the set of discrete temperature values ​​in sequence;

[0011] S22, at each discrete temperature value, ensures that the blackbody radiation source fills the field of view of the infrared thermal imager;

[0012] S23, at each discrete temperature value, the temperature of the blackbody radiation source is measured using an infrared thermal imager to obtain the temperature values ​​of all pixels of the infrared thermal imager; using the temperature values ​​of all pixels, a subset of measured temperatures corresponding to the discrete temperature values ​​is constructed.

[0013] S24, using the temperature values ​​of all pixels of the infrared thermal imager obtained at all discrete temperature values ​​of the blackbody radiation source, a set of measurement temperatures is constructed; the set of measurement temperatures includes a subset of measurement temperatures corresponding to each discrete temperature value.

[0014] The process of processing the measured temperature set to obtain the measurement temperature correction model for the infrared thermal imager includes:

[0015] S31, preprocess the measured temperature set to obtain the temperature difference set;

[0016] S32, construct a correction model for the temperature difference set to obtain the measurement temperature correction model of the infrared thermal imager.

[0017] The preprocessing of the measured temperature set to obtain the temperature difference set includes:

[0018] S311, for each element of each subset of measured temperatures in the set of measured temperatures, subtract the corresponding discrete temperature value to obtain the corresponding subset of difference temperatures;

[0019] S312, For each subset of temperature differences, perform classification and discrimination processing to obtain the corresponding subset of temperature differences; each subset of temperature differences has a corresponding discrete temperature value;

[0020] S313, using all the temperature difference subsets, construct the temperature difference set.

[0021] The process of classifying and discriminating each subset of temperature differences to obtain the corresponding subset of temperature differences includes:

[0022] S3121, using the difference temperature value of each pixel in each difference temperature subset and the two-dimensional coordinates of the pixel, a three-dimensional space data point is constructed, and using all the pixels of the difference temperature subset, a three-dimensional space data point set is constructed.

[0023] S3122, perform cluster analysis on the data point set to obtain category information and the data points contained in each category;

[0024] S3123, for each category of information, calculate the corresponding central data point;

[0025] S3124, For each category of information, calculate the distance between the data points and the center data point; determine whether the distance is greater than a preset distance threshold, and obtain a first discrimination result;

[0026] S3125, delete the data points whose first discrimination result is greater than the category information;

[0027] S3126, Perform S3123 to S3125 on all category information to obtain the data points contained in the discriminated category information;

[0028] S3127, using the data points contained in all the classified category information, a temperature difference subset is constructed.

[0029] The calculation expression for the central data point is:

[0030] ,

[0031] ,

[0032] ,

[0033] in, Let N be the value of the center data point, and N be the total number of data points included in the category information. , and This is the mean calculated for all data points included in the category information. The value of the i-th data point contained in the category information.

[0034] The step of constructing a correction model for the temperature difference set to obtain the measurement temperature correction model for the infrared thermal imager includes:

[0035] S321, using the two-dimensional coordinates of the pixels in the temperature difference subset corresponding to each discrete temperature value as ternary independent variables, and the difference temperature value of the pixels in the temperature difference subset corresponding to the discrete temperature value as the dependent variable, multiple regression modeling is performed on all ternary independent and dependent variables to obtain a set of regression functions; the set of regression functions includes a first regression function, a second regression function, and a third regression function; the regression methods used for the first regression function, the second regression function, and the third regression function are least squares regression, structural equation model (SEM) regression, and ridge regression, respectively;

[0036] S322, Perform performance testing on the set of regression functions to obtain a set of performance indicators for each regression function; the set of performance indicators includes mean squared error, F-value, and standard deviation of SHAP value;

[0037] S323, based on all performance index sets, performs fusion calculations on the regression function set to obtain the measurement temperature correction model for the infrared thermal imager.

[0038] A second aspect of this invention discloses a temperature measurement error estimation device for an infrared thermal imager, the device comprising:

[0039] Memory containing executable program code;

[0040] A processor coupled to the memory;

[0041] The processor calls the executable program code stored in the memory to execute the infrared thermal imager temperature measurement error estimation method.

[0042] In a third aspect of this invention, a computer-readable storage medium is disclosed, wherein the computer-readable storage medium stores computer instructions, which, when invoked by a computer, are used to execute the temperature measurement error estimation method of the infrared thermal imager.

[0043] In a fourth aspect of this invention, an information data processing terminal is disclosed, which is used to implement the temperature measurement error estimation method of the infrared thermal imager.

[0044] The beneficial effects of this invention are as follows:

[0045] This invention sets a discrete set of temperature values ​​and places the blackbody radiation source at each temperature sequentially. At the same time, it ensures that the blackbody radiation source fills the field of view of the infrared thermal imager for full pixel measurement. This enables the acquisition of complete measurement data covering multiple temperature ranges and the entire field of view, providing a comprehensive and reliable data foundation for subsequent error correction and effectively avoiding correction deviations caused by incomplete measurement data.

[0046] This invention performs three-dimensional spatial clustering analysis on the differential temperature subset constructed from measurement data. By combining a specially designed method for calculating the center data point and a weighted distance discrimination criterion to remove outlier data points, it can accurately screen out effective error data, reduce the interference of outliers on error estimation, and significantly improve the accuracy and effectiveness of the temperature difference set.

[0047] This invention employs three different types of regression methods to construct a set of regression functions, and introduces multi-dimensional performance indicators such as mean squared error, standard deviation of SHAP value, and F-value for comprehensive evaluation. This fully leverages the advantages of different regression algorithms, avoids the limitations of a single regression method in complex error scenarios, and enhances the adaptability of the calibration model to different error sources.

[0048] This invention integrates multiple regression functions through a specific fusion calculation method and dynamically adjusts the weights of each regression function in combination with performance indicators. This enables the constructed temperature measurement correction model to combine the advantages of each regression function, resulting in higher accuracy and robustness in temperature measurement error compensation. It can effectively reduce the temperature measurement error of infrared thermal imagers in different temperature ranges and field of view areas.

[0049] The temperature measurement error estimation method of this invention has a clear process and is highly operable. It does not require complex hardware modifications and can achieve high-precision error correction simply through data processing and model building. It is suitable for performance optimization of various infrared thermal imagers and significantly broadens the application range of infrared thermal imagers in high-precision temperature measurement scenarios. It has strong practicality and promotional value. Attached Figure Description

[0050] Figure 1 This is a flowchart illustrating the implementation of the method of the present invention. Detailed Implementation

[0051] To better understand the content of this invention, an embodiment is provided here.

[0052] Figure 1 This is a flowchart illustrating the implementation of the method of the present invention.

[0053] In a first aspect, this invention discloses a method for estimating the temperature measurement error of an infrared thermal imager, comprising:

[0054] S1 sets the discrete temperature value set of the blackbody radiation source;

[0055] S2, Based on the set of discrete temperature values, use an infrared thermal imager to measure the temperature of the blackbody radiation source to obtain a set of measured temperatures;

[0056] S3, process the measured temperature set to obtain the measurement temperature correction model of the infrared thermal imager.

[0057] The method of measuring the temperature of a blackbody radiation source using an infrared thermal imager based on the discrete temperature value set to obtain a set of measured temperatures includes:

[0058] The set of discrete temperature values ​​includes discrete temperature values ​​of several blackbody radiation sources.

[0059] S21, so that the blackbody radiation source is sequentially placed under each discrete temperature value in the set of discrete temperature values;

[0060] S22, at each discrete temperature value, ensures that the blackbody radiation source fills the field of view of the infrared thermal imager;

[0061] S23, at each discrete temperature value, the temperature of the blackbody radiation source is measured using an infrared thermal imager to obtain the temperature values ​​of all pixels of the infrared thermal imager; using the temperature values ​​of all pixels, a subset of measured temperatures corresponding to the discrete temperature values ​​is constructed.

[0062] S24, using the temperature values ​​of all pixels of the infrared thermal imager obtained at all discrete temperature values ​​of the blackbody radiation source, a set of measurement temperatures is constructed; the set of measurement temperatures includes a subset of measurement temperatures corresponding to each discrete temperature value.

[0063] The process of processing the measured temperature set to obtain the measurement temperature correction model for the infrared thermal imager includes:

[0064] S31, preprocess the measured temperature set to obtain the temperature difference set;

[0065] S32, construct a correction model for the temperature difference set to obtain the measurement temperature correction model of the infrared thermal imager.

[0066] The preprocessing of the measured temperature set to obtain the temperature difference set includes:

[0067] S311, for each element of each subset of measured temperatures in the set of measured temperatures, subtract the corresponding discrete temperature value to obtain the corresponding subset of difference temperatures;

[0068] S312, For each subset of temperature differences, perform classification and discrimination processing to obtain the corresponding subset of temperature differences; each subset of temperature differences has a corresponding discrete temperature value;

[0069] S313, using all the temperature difference subsets, construct the temperature difference set.

[0070] The process of classifying and discriminating each subset of temperature differences to obtain the corresponding subset of temperature differences includes:

[0071] S3121, using the difference temperature value of each pixel in each difference temperature subset and the two-dimensional coordinates of the pixel, a three-dimensional space data point is constructed, and using all the pixels of the difference temperature subset, a three-dimensional space data point set is constructed.

[0072] The elements of a data point, including the temperature difference value of each pixel and the two-dimensional coordinates of the pixel, can be expressed in the form of [x,y,z].

[0073] S3122, perform cluster analysis on the data point set to obtain category information and the data points contained in each category;

[0074] S3123, for each category of information, calculate the corresponding central data point;

[0075] S3124, For each category of information, calculate the distance between the data points and the center data point; determine whether the distance is greater than a preset distance threshold, and obtain a first discrimination result;

[0076] S3125, delete the data points whose first discrimination result is greater than the category information;

[0077] S3126, Perform S3123 to S3125 on all category information to obtain the data points contained in the discriminated category information;

[0078] S3127, using the data points contained in all the classified category information, a temperature difference subset is constructed.

[0079] The calculation expression for the central data point is:

[0080] ,

[0081] ,

[0082] ,

[0083] in, Let N be the value of the center data point, and N be the total number of data points included in the category information. , and This is the mean calculated for all data points included in the category information. The value of the i-th data point contained in the category information.

[0084] The central data point calculation expression of this invention combines the statistical mean characteristics of all data points within a category, dynamically weighting each data point through the synergistic effect of trigonometric and exponential functions. This design effectively weakens the interference of individual extreme data points on the central position, prevents the central data point from shifting towards outliers, and ensures that the calculated central data point accurately reflects the core distribution characteristics of the corresponding category of data. Simultaneously, the expression fully considers the attributes of data points in each dimension of three-dimensional space, enabling the central data point to comprehensively conform to the aggregation patterns of real data. This provides a reliable and accurate benchmark for subsequent outlier identification, ensuring that the subsequently selected valid data accurately reflects the actual error distribution of the infrared thermal imager.

[0085] The expression for calculating the distance between the data point and the center data point is:

[0086]

[0087] in, and These are preset weighting factors, with values ​​of 0.4 and 0.6 respectively. and These represent the vectors corresponding to the data points and the center data point, respectively. This indicates that the calculation yielded the result. Euclidean norm, This indicates that the calculation yielded the result. The Frobenius norm of the vector, where the elements are the three element values ​​of the data points;

[0088] The step of constructing a correction model for the temperature difference set to obtain the measurement temperature correction model for the infrared thermal imager includes:

[0089] S321, using the two-dimensional coordinates of the pixels in the temperature difference subset corresponding to each discrete temperature value as ternary independent variables, and the difference temperature value of the pixels in the temperature difference subset corresponding to the discrete temperature value as the dependent variable, multiple regression modeling is performed on all ternary independent and dependent variables to obtain a set of regression functions; the set of regression functions includes a first regression function, a second regression function, and a third regression function; the regression methods used for the first regression function, the second regression function, and the third regression function are least squares regression, structural equation model (SEM) regression, and ridge regression, respectively;

[0090] S322, Perform performance testing on the set of regression functions to obtain a set of performance indicators for each regression function; the set of performance indicators includes mean squared error, standard deviation of SHAP value, and F-value;

[0091] The performance test can be performed using hypothesis testing methods;

[0092] S323, based on all performance index sets, performs fusion calculations on the regression function set to obtain the measurement temperature correction model for the infrared thermal imager.

[0093] The expression for the fusion calculation is:

[0094] ,

[0095] in, This represents the measurement temperature correction model for an infrared thermal imager. Let j represent the regression function. , , Let represent the mean squared error, standard deviation of the SHAP value, and F-value of the j-th regression function, respectively. and Let F represent the maximum value of all mean squared errors and the maximum value of all F-values, respectively. is a preset adjustment factor, which is a constant.

[0096] The fusion calculation expression of this invention fully utilizes the multi-dimensional performance indicators of each regression function, and achieves differentiated weighted fusion of different regression functions through the dynamic adjustment of trigonometric and exponential functions. This expression can automatically adjust the weight ratio according to the performance of each regression function, highlighting the core role of regression functions with small mean square error, strong stability, and excellent fitting effect, while also fully considering the complementary advantages of different regression algorithms in error compensation scenarios, avoiding the limitations of single regression models or simple weighted fusion. Simultaneously, by introducing adjustment factors to optimize the fusion process, the rationality and stability of the fusion results are further improved, enabling the final constructed temperature measurement correction model to comprehensively absorb the advantages of various regression functions, significantly improving the adaptability and compensation accuracy for complex temperature measurement errors, and ensuring that infrared thermal imagers can achieve high-precision temperature measurement error correction in different temperature ranges and different fields of view.

[0097] In all embodiments of the present invention, the variables involved in all computational expressions or mathematical functions have been dimensionlessized before computation.

[0098] In all embodiments of the present invention, the values ​​of the independent variables in the input of all computational expressions or mathematical functions meet the reasonable requirements of the input range of the computational expressions or mathematical functions, and can ensure that the computational expressions or mathematical functions can be calculated smoothly without violating physical laws or mathematical rules.

[0099] A second aspect of this invention discloses a temperature measurement error estimation device for an infrared thermal imager, the device comprising:

[0100] Memory containing executable program code;

[0101] A processor coupled to the memory;

[0102] The processor calls the executable program code stored in the memory to execute the infrared thermal imager temperature measurement error estimation method.

[0103] In a third aspect of this invention, a computer-readable storage medium is disclosed, wherein the computer-readable storage medium stores computer instructions, which, when invoked by a computer, are used to execute the temperature measurement error estimation method of the infrared thermal imager.

[0104] In a fourth aspect of this invention, an information data processing terminal is disclosed, which is used to implement the temperature measurement error estimation method of the infrared thermal imager.

[0105] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A method for estimating the temperature measurement error of an infrared thermal imager, characterized in that, include: S1, Set a set of discrete temperature values ​​for a blackbody radiation source; the set of discrete temperature values ​​includes discrete temperature values ​​of several blackbody radiation sources. S2, Based on the set of discrete temperature values, use an infrared thermal imager to measure the temperature of the blackbody radiation source to obtain a set of measured temperatures; S3, Process the measured temperature set to obtain the measurement temperature correction model of the infrared thermal imager, including: S31, preprocess the measured temperature set to obtain a temperature difference set, including: S311, for each element of each subset of measured temperatures in the set of measured temperatures, subtract the corresponding discrete temperature value to obtain the corresponding subset of difference temperatures; S312, for each subset of temperature differences, perform classification and discrimination processing to obtain the corresponding subset of temperature differences; each subset of temperature differences has a corresponding discrete temperature value, including: S3121, using the difference temperature value of each pixel in each difference temperature subset and the two-dimensional coordinates of the pixel, a three-dimensional space data point is constructed, and using all the pixels of the difference temperature subset, a three-dimensional space data point set is constructed. S3122, perform cluster analysis on the data point set to obtain category information and the data points contained in each category; S3123, for each category of information, calculate the corresponding central data point; The calculation expression for the central data point is: ， ， ， in, Let N be the value of the center data point, and N be the total number of data points included in the category information. , and This is the mean calculated for all data points included in the category information. The value of the i-th data point contained in the category information; S3124, For each category of information, calculate the distance between the data points and the center data point; determine whether the distance is greater than a preset distance threshold, and obtain a first discrimination result; S3125, delete the data points whose first discrimination result is greater than the category information; S3126, Perform S3123 to S3125 on all category information to obtain the data points contained in the discriminated category information; S3127, using the data points contained in all the classified category information, a subset of temperature differences is constructed; S313, using all the subsets of temperature differences, construct the set of temperature differences; S32, construct a correction model for the temperature difference set to obtain the measurement temperature correction model for the infrared thermal imager, including: S321, using the two-dimensional coordinates of the pixels in the temperature difference subset corresponding to each discrete temperature value as ternary independent variables, and the difference temperature value of the pixels in the temperature difference subset corresponding to the discrete temperature value as the dependent variable, multiple regression modeling is performed on all ternary independent and dependent variables to obtain a set of regression functions; the set of regression functions includes a first regression function, a second regression function, and a third regression function; the regression methods used for the first regression function, the second regression function, and the third regression function are least squares regression, structural equation model (SEM) regression, and ridge regression, respectively; S322, Perform performance testing on the set of regression functions to obtain a set of performance indicators for each regression function; the set of performance indicators includes mean squared error, F-value, and standard deviation of SHAP value; S323, based on all performance index sets, performs fusion calculations on the regression function set to obtain the measurement temperature correction model for the infrared thermal imager; The expression for the fusion calculation is: ， in, This represents the measurement temperature correction model for an infrared thermal imager. Let j represent the regression function. , , Let represent the mean squared error, standard deviation of the SHAP value, and F-value of the j-th regression function, respectively. and Let F represent the maximum value of all mean squared errors and the maximum value of all F-values, respectively. is a preset adjustment factor, which is a constant.

2. The method for estimating the temperature measurement error of an infrared thermal imager as described in claim 1, characterized in that, The method of measuring the temperature of a blackbody radiation source using an infrared thermal imager based on the discrete temperature value set to obtain a set of measured temperatures includes: S21, Set up a blackbody radiation source, which is placed under each discrete temperature value in the set of discrete temperature values ​​in sequence; S22, at each discrete temperature value, ensures that the blackbody radiation source fills the field of view of the infrared thermal imager; S23, at each discrete temperature value, the temperature of the blackbody radiation source is measured using an infrared thermal imager to obtain the temperature values ​​of all pixels of the infrared thermal imager; using the temperature values ​​of all pixels, a subset of measured temperatures corresponding to the discrete temperature values ​​is constructed. S24, using the temperature values ​​of all pixels of the infrared thermal imager obtained at all discrete temperature values ​​of the blackbody radiation source, a set of measurement temperatures is constructed; the set of measurement temperatures includes a subset of measurement temperatures corresponding to each discrete temperature value.

3. A temperature measurement error estimation device for an infrared thermal imager, characterized in that, The device includes: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the temperature measurement error estimation method of the infrared thermal imager as described in any one of claims 1 to 2.

4. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, which, when invoked by a computer, are used to execute the temperature measurement error estimation method of the infrared thermal imager as described in any one of claims 1 to 2.

5. An information data processing terminal, characterized in that, The information data processing terminal is used to implement the temperature measurement error estimation method of the infrared thermal imager as described in any one of claims 1 to 2.

Images