A medium-long wave infrared imaging spectrometer relative radiation calibration system and method

Abstract

The application discloses a kind of middle long wave infrared imaging spectrometer relative radiation calibration system and method, belong to space infrared imaging technical field, solve middle long wave infrared imaging spectrometer segmented two-point method relative radiation calibration exists response reference value is susceptible to bad pixel influence, pixel mapping dynamic range is small, lack of continuity at temperature interval junction technical problem, its relative radiation calibration system includes blackbody radiation source, blackbody temperature controller, exhaust platform, data acquisition processing system and clean laboratory.Blackbody radiation source simulates the target radiation of different temperature, located in the field of view center of infrared imaging spectrometer, exhaust platform provides on-board environment.Its method includes detector dynamic range division, data image acquisition and average, background noise removal, blackbody brightness calculation, linear relationship coefficient is obtained, response reference point is determined, reference linear relationship coefficient, relative radiation correction coefficient is obtained.The application is used for target material composition, temperature distribution and surface characteristic identification.

Description

Technical Field

[0001] This invention belongs to the field of space infrared imaging technology, specifically relating to a relative radiometric calibration system and method for a mid-to-long-wave infrared imaging spectrometer. Background Technology

[0002] Mid- and long-wave infrared high-resolution spectroscopic imagers are advanced optical instruments based on infrared spectroscopy technology. They can acquire continuous spectral and spatial information of targets in the mid-infrared (3-5 micrometers) and long-wave infrared (8-14 micrometers) bands of the electromagnetic spectrum. Their core principle is to detect the radiation characteristics of target objects in the infrared band, combining spectral analysis and imaging technology to achieve high-precision identification and analysis of the target's material composition, temperature distribution, and surface characteristics. They are widely used in environmental monitoring, military reconnaissance, industrial inspection, medical diagnosis, and astronomical observation. Especially in the military and security fields, mid- and long-wave infrared imaging spectrometers, with their excellent performance at night and in adverse weather conditions, are used for target detection, concealed target identification, and long-range detection of chemical warfare agents. However, their quantitative application still faces several bottlenecks: first, the cooling requirements and high noise levels of infrared detectors limit data accuracy and real-time performance; second, calibrating radiative transfer models and eliminating background interference in complex environments is difficult; and third, the processing and analysis algorithms for spectral data need further optimization to achieve efficient and accurate information extraction. Therefore, high-precision relative radiometric calibration is a crucial guarantee for the accurate application of this type of instrument.

[0003] Detector inhomogeneity refers to the difference in response of detector pixels to the same radiation input in an infrared imaging spectrometer. This is mainly manifested in inconsistent gain and offset between pixels. This inhomogeneity introduces fixed-mode noise (FPN), causing stripes or spots in the image, severely affecting data quality and the accuracy of quantitative analysis. The core significance of relative radiometric calibration lies in eliminating the influence of detector inhomogeneity, optical system attenuation, and environmental interference by calibrating the instrument's response function. This converts the raw digital signal into physically meaningful radiance values, ensuring data consistency and comparability. This process not only provides a reliable data foundation for quantitative applications such as temperature inversion, component analysis, and substance identification, but also provides a standardized framework for multi-source data fusion, long-term monitoring, and cross-platform comparative studies. It is the core technical support for the scientific value and application effectiveness of hyperspectral imaging data.

[0004] Currently, the pre-emission relative radiometric calibration of mid- and long-wave infrared imaging spectrometers mainly employs laboratory calibration methods based on blackbody radiation sources. This involves simulating target radiation at different temperatures using a high-precision blackbody radiation source in a controlled environment, combining this with the response characteristics of a standard detector to establish a pixel-level radiometric response function (including gain and offset coefficients), and using a uniform radiation field to correct detector non-uniformity. Furthermore, the calibration process must consider the effects of optical system attenuation, stray light effects, and ambient temperature variations. Multi-temperature point calibration and radiative transfer model optimization ensure the accuracy of calibration data and mitigate errors introduced by the nonlinearity of the instrument response. However, current mainstream radiometric calibration methods primarily use the two-point method and the piecewise two-point method to calculate the relative radiometric calibration coefficients of infrared detectors. The two-point method uses a linear model to calculate the relative radiometric calibration coefficients, which does not account for the nonlinearity of the infrared detector response, resulting in reduced calibration accuracy at the low and high ends of the infrared detector response curve. The piecewise two-point method avoids the influence of the nonlinear response of the infrared detector. The mainstream piecewise two-point method uses the row mean of the infrared detector response as the reference value for calculating the relative radiometric calibration coefficient. This method is easily affected by bad pixels, resulting in unstable reference values ​​and inaccurate relative radiometric calibration coefficients. In addition, the mainstream piecewise two-point method uses the row mean of the infrared detector response as the threshold for the use of relative radiometric calibration coefficients in various temperature ranges. This method will cause the response curves of some pixels with small dynamic ranges to be mapped only to the high or low end of the reference response curve, resulting in very limited correction effect of the calculated relative radiometric calibration coefficients and low relative radiometric calibration accuracy. At the junction of temperature ranges, the mainstream piecewise two-point method may cause a step in the corrected image data due to accidental piecewise fluctuations, resulting in a lack of continuity between temperature ranges. Summary of the Invention

[0005] To overcome the shortcomings of the two-point segmented relative radiometric calibration method for mid- and long-wave infrared imaging spectrometers, such as the susceptibility of the response reference value to bad pixels, the small dynamic range of pixel mapping, the poor relative radiometric correction effect, and the lack of continuity at the junction of temperature ranges, this invention proposes a relative radiometric calibration system and method for mid- and long-wave infrared imaging spectrometers.

[0006] The technical solution adopted by this invention to solve its technical problem is:

[0007] A relative radiometric calibration system for a medium- and long-wave infrared imaging spectrometer includes a blackbody radiation source, a blackbody temperature controller, an exhaust platform, a data acquisition and processing system, and a clean laboratory.

[0008] The blackbody radiation source, blackbody temperature controller, exhaust platform, data acquisition and processing system, and infrared imaging spectrometer are located in the clean laboratory.

[0009] The blackbody radiation source is used to simulate target radiation at different temperatures, the blackbody temperature controller is used to control the temperature of the blackbody radiation source, the exhaust platform and detector cooling system are used to provide the on-board environment, the data acquisition and processing system is used to synchronously record the output data of the medium and long-wave infrared imaging spectrometer, and the clean laboratory is used to provide a clean environment.

[0010] The blackbody radiation source is placed at the center of the field of view of the infrared imaging spectrometer, filling the field of view; the detector of the infrared imaging spectrometer acquires data images of the blackbody radiation source, and the data images are processed by the data acquisition and processing system.

[0011] An infrared imaging spectrometer includes a detector and a detector cooling system.

[0012] In the aforementioned medium- and long-wave infrared imaging spectrometer relative radiometric calibration system, the blackbody radiation source has a uniform temperature.

[0013] In the aforementioned medium- and long-wave infrared imaging spectrometer relative radiometric calibration system, the blackbody radiation source has a temperature of 200K~350K and an emissivity ≥0.995.

[0014] The blackbody temperature controller has an accuracy of ±0.01K, the exhaust pressure of the exhaust platform is ≤1×10^-3 Pa, the cooling temperature of the detector cooling system is 80±0.1K, and the cleanroom is ISO Class 5, ±0.5℃.

[0015] A method for relative radiometric calibration of a mid-to-long-wave infrared imaging spectrometer includes the following steps:

[0016] Step 1, Detector dynamic range division

[0017] Based on the detector response characteristics of the infrared imaging spectrometer, the detector dynamic range is divided into several intervals, with a temperature difference of ≥10℃ at the endpoints of each interval. The temperature at the endpoint of each interval is the output temperature of the blackbody radiation source.

[0018] Step 2, Data Image Acquisition and Averaging

[0019] At the temperature endpoint of each interval, the detector acquires data images of the blackbody radiation source, and performs time-domain averaging on the acquired data images to obtain stable data images of the endpoint temperatures.

[0020] Step 3, Background Noise Removal

[0021] The temperature of the blackbody radiation source is set to 100K~180K. The detector collects dark background data images of the blackbody radiation source and performs time-domain averaging on the collected dark background data images to obtain stable dark background data images.

[0022] Subtracting the stable data image of the dark background from the stable data image of the endpoint temperature yields the data image of the endpoint temperature after removing background noise.

[0023] Step 4, Blackbody Radiance Calculation

[0024] Using the Planck function, calculate the blackbody radiance of the blackbody radiation source at the temperature at each endpoint of the interval.

[0025] Step 5: Calculate the linear relationship coefficients.

[0026] In each temperature range, by combining the blackbody radiance and the pixel response data of the background-noise-removed data image with the endpoint temperature, a linear relationship between the two is established, and the linear relationship coefficient between the blackbody radiance and the pixel response data of the background-noise-removed data image is obtained.

[0027] Step 6, Determine the response reference point

[0028] The background noise-removed data of the endpoint temperature data image is sorted by different spectral bands. The median pixel of each spectrum at each temperature point is found. The pixel that is the median at each temperature point is selected as the response reference point.

[0029] Step 7: Calculate the reference linear relationship coefficient.

[0030] In each temperature range, by combining the blackbody radiance and the response data of each pixel at the response reference point, a linear relationship between the two is established, resulting in a reference linear relationship coefficient between the blackbody radiance and the pixel response data of the response reference point.

[0031] Step 8: Calculate the relative radiation correction factor

[0032] By using linear relationship coefficients and reference linear relationship coefficients, and through linear coefficient transformation, the relative radiation correction coefficients for each pixel in different temperature ranges are obtained.

[0033] In the above-mentioned relative radiometric calibration method for medium and long-wave infrared imaging spectrometers, step 1 involves 5-10 temperature intervals.

[0034] The aforementioned relative radiometric calibration method for mid- and long-wave infrared imaging spectrometers also includes:

[0035] Step 9, Select to use a threshold

[0036] For each temperature range, the response data of each pixel in the background noise-removed image is selected as the range division threshold for the relative radiometric correction coefficient.

[0037] Step 10, Image Response Value Correction

[0038] Using the relative radiation correction coefficient and the threshold value of the relative radiation correction coefficient, a weighted fusion algorithm is employed to correct the data image acquired by the detector, thus obtaining the corrected data image.

[0039] The aforementioned relative radiometric calibration method for mid- and long-wave infrared imaging spectrometers, further includes step 2:

[0040] Endpoint temperature The response value of each pixel in the image is obtained by averaging the stable data from multiple frames. The calculation formula is as follows:

[0041]

[0042] In the formula, The original data, These are the temperatures at the two ends of the temperature range. , The total number of endpoint temperatures, temperature range The endpoint temperature is , , , The row number of the cell. , This represents the total number of rows in the data image. The column number of the cell. , This represents the total number of columns in the data image. For data frame sequence number, Temperature point The total number of frames corresponding to the data;

[0043] Repeat the above steps to obtain Pixel response data for each endpoint temperature: , , ..., , ;

[0044] Step 3 further includes:

[0045] Acquire dark background data images at temperatures ranging from 100K to 180K. ;

[0046] Response value of each pixel in a data image after background noise removal The calculation formula is as follows:

[0047]

[0048] In step 4, the blackbody radiance The calculation formula is as follows:

[0049]

[0050] In the above formula, For the first The center wavelength corresponding to a row pixel is related to the characteristics of the infrared spectral imager and can be obtained from the calibration report. is Planck's constant. , At the speed of light, , Boltzmann's constant, ;

[0051] Step 5 further includes:

[0052] No. The linear relationship coefficients for each temperature range include the linear gain coefficient. and linear offset coefficient The calculation formula is as follows:

[0053]

[0054]

[0055] In the above formula, This refers to the temperature range number. ;

[0056] Step 6 further includes:

[0057] The spectral data for different temperature bands are sorted, and the median pixel for each spectrum at each temperature point is found. After filtering, the pixels that are the median at every temperature point are used as the response reference points. The pixel response value at each temperature is used as a response reference value. .

[0058] Step 7 further includes:

[0059] No. The linear relationship coefficients of the response reference points for each temperature range include the linear gain coefficient. and linear offset coefficient The calculation formula is as follows:

[0060]

[0061]

[0062] Step 8 further includes:

[0063] The relative radiation correction factor includes the correction gain factor. Correction offset coefficient The calculation formula is as follows:

[0064]

[0065]

[0066] The aforementioned method for relative radiometric calibration of mid- and long-wave infrared imaging spectrometers, step 9, further includes:

[0067] Use threshold The calculation formula is as follows:

[0068]

[0069] Step 10 further includes:

[0070] The response data of each pixel in the corrected image. The calculation formula is as follows:

[0071]

[0072] In the above formula, The coordinate position is The pixels in the temperature range The correction value, , The coordinates of the infrared image to be corrected are: The pixel response value, The coordinate position is The pixels in the temperature range The correction value, , The coordinate position is The pixels in the temperature range The weight value, , The coordinate position is The pixels in the temperature range The weight value, .

[0073] The beneficial effects of this invention are:

[0074] A relative radiometric calibration method for a mid-to-long-wave infrared imaging spectrometer effectively compensates for the nonlinear response of the detector through local linear approximation, and ensures calibration continuity by combining it with an inter-segment weighted fusion algorithm.

[0075] A relative radiometric calibration method for a medium- and long-wave infrared imaging spectrometer is proposed. This method uses the median response of each temperature point in each spectral channel as the reference response to calculate the reference linear relationship coefficient. This method can greatly reduce the impact of outliers and has extremely high robustness.

[0076] A relative radiometric calibration method for mid-to-long-wave infrared imaging spectrometers innovatively selects a temperature-divided response value range for each pixel based on the inconsistency of the response of each detector pixel. Specifically, it uses the response values ​​of each pixel at different temperatures as the threshold for dividing the application range of the relative radiometric correction coefficient, thus avoiding situations where the response of some pixels can only be mapped to the low or high end of the reference pixel response. This improves the accuracy of relative radiometric correction. Attached Figure Description

[0077] Figure 1 This is a schematic diagram of the relative radiometric calibration system of the long-wave infrared imaging spectrometer in Embodiment 2 of the present invention;

[0078] Figure 2 This is a flowchart of the relative radiometric calibration method for a long-wave infrared imaging spectrometer in Embodiment 4 of the present invention;

[0079] Figure 3 This describes the process of selecting reference points for the response of different wavebands at various temperature points in Embodiment 4 of the present invention.

[0080] Figure 4 This is a comparison diagram of the mapping between using reference cell response partitioning and using single-point cell response partitioning in Embodiment 4 of the present invention. Figure 4 (a) is a mapping diagram using the reference cell response row mean partitioning. Figure 4 (b) is a mapping diagram using single-cell response partitioning;

[0081] Figure 5 This is a comparison of the imaging images of the blackbody radiation source in Embodiment 4 of the present invention at a temperature of 280K before and after relative radiation correction. Figure 5 (a) is the infrared image to be corrected before relative radiometric correction. Figure 5 (b) is the infrared image after relative radiometric correction;

[0082] Figure 6 This is a comparison diagram of the spatial dimensional pixel DN value changes of the blackbody radiation source in Embodiment 4 of the present invention at a temperature of 280K before and after relative radiation correction. Figure 6 (a) represents the spatial dimension pixel DN value of the infrared image to be corrected before relative radiometric correction. Figure 6 (b) represents the spatial dimension pixel DN value of the infrared image after relative radiometric correction.

[0083] The attached figures are labeled as follows:

[0084] 1. Blackbody radiation source, 2. Blackbody temperature controller, 3. Exhaust platform, 4. Detector cooling system, 5. Infrared imaging spectrometer, 6. Data acquisition and processing system, 7. Clean laboratory. Detailed Implementation

[0085] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0086] Example 1

[0087] An improved relative radiometric calibration technique for a segmented two-point mid-to-long-wave infrared high-resolution spectral imager includes the following steps:

[0088] 1. An improved relative radiometric calibration technique for a segmented two-point mid-to-long-wave infrared high-resolution spectral imager, characterized by comprising the following steps:

[0089] S1 covers the entire dynamic range of the detector and is divided into N intervals, with two temperature points selected in each interval.

[0090] The temperature range is divided to cover the full dynamic range of the detector.

[0091] S2, at different temperature points, simultaneously acquires multiple frames of images from the detector of the high-resolution infrared imaging spectrometer and performs time-domain averaging to suppress random noise.

[0092] The detector's multi-temperature point response data was obtained by averaging multiple frames and eliminating random noise.

[0093] S3: Acquire dark background data images of the blackbody radiation source at temperatures ranging from 100K to 180K, and obtain stable dark background data after averaging across multiple frames. Subtract the averaged dark background data from the stable multi-temperature data to remove background noise.

[0094] By subtracting the dark background after averaging multiple frames, the influence of background noise on relative radiometric calibration is removed.

[0095] S4 uses the Planck function to calculate the blackbody radiance of the blackbody radiation source at various temperatures.

[0096] S5, in each temperature range, calculates the linear relationship coefficient between the blackbody radiance and the response data of each pixel of the detector.

[0097] S6. Sort the data of different spectral bands at each temperature point, find the median pixel of each spectrum at each temperature point, and filter to obtain the pixel that is the median at each temperature point as the response reference point. The pixel response value of the pixel at each temperature is used as the response reference value.

[0098] Using the median point of the common pixel response in different spectral bands at different temperatures as a reference point has strong robustness and can reduce the influence of outliers on the reference response.

[0099] S7. Combining the blackbody radiance and the response data of the response reference point, the reference linear relationship coefficient between the reference response value and the blackbody radiance is obtained.

[0100] By using the Planck function to calculate the linear relationship coefficients between the response of the pixel to be calibrated, the response of the reference pixel, and the blackbody radiance in different temperature ranges, the influence of detector response nonlinearity can be reduced.

[0101] S8. After obtaining the reference linear relationship coefficients for each spectral band in each temperature range and the linear relationship coefficients for each pixel in each spectral band, the correction coefficients for each pixel to be corrected to the response of the reference pixel are obtained through linear coefficient conversion, i.e., the relative radiation correction coefficients.

[0102] S9, for each temperature range, use the response value of each pixel at the current temperature as the range division threshold for the relative radiation correction coefficient.

[0103] For each temperature range, based on the inconsistency of the response of each detector pixel, an innovative approach is taken to establish a set of temperature-divided response value ranges for each pixel. That is, the response values ​​of each pixel at different temperatures are used as the threshold for dividing the range of relative radiometric correction coefficients. This feature avoids high-end and low-end mapping of each pixel in the process of mapping the response of the reference pixel, thereby improving the accuracy of relative radiometric correction.

[0104] S10 After obtaining the segmented relative radiometric correction coefficient for each pixel, the infrared image to be corrected can be divided into temperature ranges according to the reference value corresponding to each pixel in each temperature range, and the image response value can be corrected according to the correction coefficient of the corresponding range to remove non-uniform noise.

[0105] At the junctions of different temperature ranges, a weighted fusion algorithm between segments is used to enhance the continuity of the partitions.

[0106] Example 2

[0107] A pre-launch relative radiometric calibration system for a high-resolution space infrared imaging spectrometer with a spatial resolution of 3m or higher is provided. The system includes a high-precision blackbody radiation source 1, a blackbody temperature controller 2, a high-vacuum exhaust platform 3, an infrared detector cooling system 4, the high-resolution infrared imaging spectrometer under test 5, a data acquisition and processing system 6, and a cleanroom 7. Figure 1 The high-precision blackbody radiation source is used to simulate target radiation at different temperatures; the blackbody temperature controller is used to control the temperature of the blackbody, ensuring its stability and uniformity; the high-vacuum exhaust platform and the infrared detector cooling system provide a vacuum low-temperature environment for the high-resolution infrared imaging spectrometer, allowing it to operate in an environment similar to that on a satellite; the data acquisition and processing system is used to synchronously record the output data of the high-resolution infrared imaging spectrometer and perform subsequent data processing such as multi-temperature point calibration; the clean laboratory provides a stable test temperature and clean environment for the entire calibration system.

[0108] This invention proposes a segmented two-point relative radiometric calibration method to address the non-uniformity of detectors in high-resolution infrared imaging spectrometers. By dividing the detector's dynamic range into several linear intervals and calibrating them independently, the radiometric measurement accuracy within the detector's dynamic range can be significantly improved. The specific implementation steps are as follows:

[0109] S1, based on the detector's response characteristics, divide the temperature range into 5-10 intervals. For each interval, select two temperature endpoints, high and low, and output the target temperature sequentially through a blackbody radiation source to divide the temperature range covering the full dynamic range of the detector.

[0110] S2, at different temperature points, simultaneously acquire multiple frames of images from the detector of the high-resolution infrared imaging spectrometer and perform time-domain averaging to suppress random noise, thereby obtaining stable multi-temperature point data images.

[0111] S3: Acquire dark background data images of the blackbody radiation source at temperatures ranging from 100K to 180K. After averaging across multiple frames, stable dark background data is obtained. The stable multi-temperature data is then subtracted from the dark background data to remove background noise.

[0112] S4 uses the Planck function to calculate the blackbody radiance of the blackbody radiation source at various temperatures.

[0113] S5. In each temperature range, by combining the blackbody radiance and the response data of each pixel of the detector, a linear relationship between the two can be established, and the linear relationship coefficient of each pixel can be obtained.

[0114] S6. Sort the data of different spectral bands at each temperature point, find the median pixel of each spectrum at each temperature point, and filter to obtain the pixel that is the median at each temperature point as the response reference point. The pixel response value of the pixel at each temperature is used as the response reference value.

[0115] S7. Combining the blackbody radiance and the response data of the response reference point, the reference linear relationship coefficient between the reference response value and the blackbody radiance is obtained.

[0116] S8. After obtaining the reference linear relationship coefficients for each spectral band in each temperature range and the linear relationship coefficients for each pixel in each spectral band, the correction coefficients for each pixel to be corrected to the response of the reference pixel are obtained through linear coefficient conversion, i.e., the relative radiation correction coefficients.

[0117] S9, for each temperature range, based on the inconsistency of the response of each detector pixel, innovatively selects to establish a set of temperature-divided response value ranges for each pixel, that is, using the response value of each pixel at different temperatures as the threshold for dividing the range of relative radiation correction coefficients.

[0118] S10 After obtaining the segmented relative radiometric correction coefficient for each pixel, the infrared image to be corrected can be divided into temperature ranges according to the reference value corresponding to each pixel in each temperature range, and the image response value can be corrected according to the correction coefficient of the corresponding range to remove non-uniform noise.

[0119] At the junctions of different temperature ranges, a weighted fusion algorithm between segments is used to enhance the continuity of the partitions.

[0120] Example 3

[0121] To establish a pre-launch relative radiometric calibration system for a high-resolution space infrared imaging spectrometer, the system includes a high-precision blackbody radiation source (200-350K, emissivity ≥0.995), a precision temperature control system (±0.01K), a high-vacuum environment simulation device (≤1×10^-3 Pa), a cryogenic refrigeration system (80±0.1K), a data acquisition and processing unit (16bit@100kS / s), and a clean laboratory (ISO Class 5, ±0.5℃).

[0122] Example 4

[0123] The high-resolution medium- and long-wave space infrared imaging spectrometer uses a multi-segment two-point relative radiometric correction algorithm before launch, combined with... Figure 2 Explanation: The piecewise two-point calibration method divides the detector's dynamic range into several linear intervals (segments), performs independent two-point calibration within each interval, and finally stitches together the calibration results of each segment. Its core idea is:

[0124] 1) Piecewise linearity assumption: The detector's response is approximately linear within a local temperature range.

[0125] 2) Interval splicing: Smooth transition between segments is ensured through interpolation or weighted fusion.

[0126] like Figure 2 As shown, the specific implementation steps of this invention are as follows:

[0127] S1, based on the detector response characteristics, divide the temperature range into 5-10 temperature ranges, and select two temperature endpoints, high and low, for each range. , The target temperature is output sequentially through a blackbody radiation source.

[0128] S2. Place the high-precision blackbody radiation source at the center of the spectrometer's field of view, ensuring it fills the field. Set the blackbody emissivity (typically 0.95~0.99, which must be matched to the spectrometer's wavelength range). Using the blackbody radiation source, at each temperature point... Multiple frames of images were acquired, and the average value was taken to suppress random noise. The response value of each pixel was recorded. The formula is as follows:

[0129]

[0130] In the formula, The original data, These are the temperatures at the two ends of the temperature range. , The total number of endpoint temperatures, temperature range The endpoint temperature is , , , The row number of the cell. , This represents the total number of rows in the data image. The column number of the cell. , This represents the total number of columns in the data image. For data frame sequence number, Temperature point The total number of frames corresponding to the data;

[0131] Repeat the above steps to obtain Pixel response data for each endpoint temperature: , , ..., , ;

[0132] S3: Acquire dark background data images when the blackbody radiation source temperature is 100K~180K, and obtain stable dark background data after averaging multiple frames. The response value of each pixel in the data image after removing background noise. The calculation formula is as follows:

[0133]

[0134] S4. Calculate the blackbody radiance of the blackbody radiation source at different wavelengths at various temperatures using the Planck function. The calculation formula is as follows:

[0135]

[0136] In the above formula, For the first The center wavelength corresponding to a row pixel is related to the characteristics of the infrared spectral imager and can be obtained from the calibration report. is Planck's constant. , At the speed of light, , Boltzmann's constant, ;

[0137] S5, in each temperature range, by combining the blackbody radiance and the response data of each pixel of the detector, a linear relationship between the two can be established, yielding the linear relationship coefficient for each pixel. The linear relationship coefficients for each temperature range include the linear gain coefficient. and linear offset coefficient The calculation formula is as follows:

[0138]

[0139]

[0140] In the above formula, This refers to the temperature range number. ;

[0141] S6, such as Figure 3 The data for different spectral bands at each temperature point are sorted, and the median pixel for each spectrum at each temperature point is found. After filtering, the pixels that are the median at every temperature point are used as the response reference points. The pixel response value at each temperature is used as a response reference value. .

[0142] S7, combining the blackbody radiance calculated using Planck's formula and the response data at the reference point, yields the linear relationship coefficient between the reference response value and the blackbody radiance in each temperature range. The linear relationship coefficients of the response reference points for each temperature range include the linear gain coefficient. and linear offset coefficient The calculation formula is as follows:

[0143]

[0144]

[0145] S8, after obtaining the reference linearity coefficients for each spectral band in each temperature range and the linearity coefficients for each pixel in each spectral band, uses linear coefficient transformation to obtain the correction coefficient for each pixel to correct to the reference pixel response, i.e., the relative radiometric correction coefficient. The relative radiometric correction coefficient includes the correction gain coefficient. Correction offset coefficient The calculation formula is as follows:

[0146]

[0147]

[0148] S9, such as Figure 4This paper compares the interval mapping using the average response row value of a reference pixel as the basis for temperature partitioning with that using the response curve of a single pixel to be corrected as the basis for temperature partitioning. It shows that using the average response row value of the reference pixel as the basis for temperature partitioning easily leads to low-end mapping and high-end mapping for low-response pixels and high-response pixels, respectively. However, using the response value of a single pixel to be corrected as the basis for temperature partitioning provides a comprehensive mapping for the entire dynamic range. Therefore, this invention innovatively selects a set of temperature-divided response value intervals for each pixel based on the inconsistency of the response of each detector pixel. That is, it uses the response value of each pixel at different temperatures as the interval partitioning threshold for the relative radiation correction coefficient, as shown in the following formula.

[0149]

[0150] S10, after obtaining the segmented relative radiometric correction coefficients for each pixel, the infrared image to be corrected can be divided into temperature ranges based on the reference values ​​corresponding to each pixel in each temperature range, and the image response values ​​can be corrected according to the correction coefficients of the corresponding ranges to remove non-uniform noise. The response data of each pixel in the corrected image... The calculation formula is as follows:

[0151]

[0152] In the above formula, The coordinate position is The pixels in the temperature range The correction value, , The coordinates of the infrared image to be corrected are: The pixel response value, The coordinate position is The pixels in the temperature range The correction value, .

[0153] like Figure 5 This is a comparison of images of a laboratory blackbody radiation source at 280K before and after relative radiation correction. Weighted averaging is used at the junctions of different temperature ranges to avoid abrupt changes. The coordinate position is The pixels in the temperature range The weight value, , The coordinate position is The pixels in the temperature range The weight value, .

[0154] Figure 6This is a comparison chart showing the change in the spatial dimensional pixel DN value of a blackbody radiation source at 280K before and after relative radiation correction. Figure 6 (a) represents the spatial dimension pixel DN value of the infrared image to be corrected before relative radiometric correction. Figure 6 (b) represents the spatial dimension pixel DN value of the infrared image after relative radiometric correction.

Claims

1. A relative radiometric calibration system for a mid-to-long-wave infrared imaging spectrometer, characterized in that, It includes a blackbody radiation source (1), a blackbody temperature controller (2), an exhaust platform (3), a data acquisition and processing system (6), and a clean laboratory (7); The blackbody radiation source (1), blackbody temperature controller (2), exhaust platform (3), data acquisition and processing system (6), and infrared imaging spectrometer (5) are located in the clean laboratory (7); The blackbody radiation source (1) is used to simulate target radiation at different temperatures. The blackbody temperature controller (2) is used to control the temperature of the blackbody radiation source (1). The exhaust platform (3) and the detector cooling system (4) are used to provide the on-board environment. The data acquisition and processing system (6) is used to synchronously record the output data of the medium and long-wave infrared imaging spectrometer. The clean laboratory (7) is used to provide a clean environment. The blackbody radiation source (1) is placed at the center of the field of view of the infrared imaging spectrometer (5) and fills the field of view; the detector of the infrared imaging spectrometer (5) collects data images of the blackbody radiation source (1), and the data images are processed by the data acquisition and processing system (6).

2. The relative radiometric calibration system for a mid-to-long-wave infrared imaging spectrometer according to claim 1, characterized in that, The blackbody radiation source (1) has a uniform temperature.

3. The relative radiometric calibration system for a mid-to-long-wave infrared imaging spectrometer according to claim 1, characterized in that, The temperature of the blackbody radiation source (1) is 200K~350K, and the emissivity is ≥0.995; The accuracy of the blackbody temperature controller (2) is ±0.01K, the exhaust pressure of the exhaust platform (3) is ≤1×10^-3 Pa, the cooling temperature of the detector cooling system (4) is 80±0.1K, and the clean laboratory (7) is ISO 5 level, ±0.5℃.

4. A method for relative radiometric calibration of a mid-to-long-wave infrared imaging spectrometer, using the relative radiometric calibration system described in any one of claims 1-3, characterized in that, Includes the following steps: Step 1, Detector dynamic range division Based on the detector response characteristics of the infrared imaging spectrometer (5), the dynamic range of the detector is divided into several intervals, with the temperature difference between the endpoints of the intervals being ≥10℃. The temperature at the endpoint of each interval is the output temperature of the blackbody radiation source (1). Step 2, Data Image Acquisition and Averaging At the end temperature of each interval, the detector acquires data images of the blackbody radiation source (1), and performs time-domain averaging on the acquired data images to obtain stable data images of the end temperature. Step 3, Background Noise Removal The temperature of the blackbody radiation source (1) is set to 100K~180K. The detector collects dark background data images of the blackbody radiation source (1). The collected dark background data images are averaged in the time domain to obtain a stable dark background data image. Subtracting the stable data image of the dark background from the stable data image of the endpoint temperature yields the data image of the endpoint temperature with background noise removed. Step 4, Blackbody Radiance Calculation Using the Planck function, calculate the blackbody radiance of the blackbody radiation source (1) at the temperature at each interval endpoint; Step 5: Calculate the linear relationship coefficients. In each temperature range, by combining the background noise-removed data image of the blackbody radiance and the endpoint temperature of each pixel, a linear relationship between the two is established, and the linear relationship coefficient between the blackbody radiance and the pixel response data of the background noise-removed data image is obtained. Step 6, Determine the response reference point The background noise-removed data images of endpoint temperatures are sorted by different spectral bands. The median pixel of each spectrum at each temperature point is found. The pixel that is the median at each temperature point is selected as the response reference point. Step 7: Calculate the reference linear relationship coefficient. In each temperature range, by combining the blackbody radiance and the response data of each pixel at the response reference point, a linear relationship between the two is established, and the reference linear relationship coefficient between the blackbody radiance and the pixel response data at the response reference point is obtained. Step 8: Calculate the relative radiation correction factor By using linear relationship coefficients and reference linear relationship coefficients, and through linear coefficient transformation, the relative radiation correction coefficients for each pixel in different temperature ranges are obtained.

5. The relative radiometric calibration method for a mid-to-long-wave infrared imaging spectrometer according to claim 4, characterized in that, In step 1, the intervals are 5-10 temperature intervals.

6. The relative radiometric calibration method for a mid-to-long-wave infrared imaging spectrometer according to claim 4 or 5, characterized in that, Also includes: Step 9, Select to use a threshold For each temperature range, the response data of each pixel in the background noise-removed image is selected as the range division threshold for the relative radiometric correction coefficient. Step 10, Image Response Value Correction Using the relative radiation correction coefficient and the threshold value of the relative radiation correction coefficient, a weighted fusion algorithm is employed to correct the data image acquired by the detector, thus obtaining the corrected data image.

7. The relative radiometric calibration method for a mid-to-long-wave infrared imaging spectrometer according to claim 4, characterized in that, Step 2 further includes: Endpoint temperature The response value of each pixel in the image is obtained by averaging the stable data from multiple frames. The calculation formula is as follows: ； In the formula, The original data, These are the temperatures at the two ends of the temperature range. , The total number of endpoint temperatures, temperature range The endpoint temperature is , , , The row number of the cell. , This represents the total number of rows in the data image. The column number of the cell. , This represents the total number of columns in the data image. For data frame sequence number, Temperature point The total number of frames corresponding to the data; Repeat the above steps to obtain Pixel response data for each endpoint temperature: , , ..., , ; Step 3 further includes: Acquire dark background data images at temperatures ranging from 100K to 180K. ; Response value of each pixel in a data image after background noise removal The calculation formula is as follows: ； In step 4, the blackbody radiance The calculation formula is as follows: ； In the above formula, For the first The center wavelength corresponding to a row pixel is related to the characteristics of the infrared spectral imager and can be obtained from the calibration report. is Planck's constant. , At the speed of light, , Boltzmann's constant, ; Step 5 further includes: No. The linear relationship coefficients for each temperature range include the linear gain coefficient. and linear offset coefficient The calculation formula is as follows: ； ； In the above formula, This refers to the temperature range number. ; Step 6 further includes: The spectral data for different temperature bands are sorted, and the median pixel for each spectrum at each temperature point is found. After filtering, the pixels that are the median at every temperature point are used as the response reference points. The pixel response value at each temperature is used as a response reference value. ; Step 7 further includes: No. The linear relationship coefficients of the response reference points for each temperature range include the linear gain coefficient. and linear offset coefficient The calculation formula is as follows: ； ； Step 8 further includes: The relative radiation correction factor includes the correction gain factor. Correction offset coefficient The calculation formula is as follows: ； 。 8. The relative radiometric calibration method for a mid-to-long-wave infrared imaging spectrometer according to claim 6, characterized in that, Step 9 further includes: Use threshold The calculation formula is as follows: ； Step 10 further includes: The response data of each pixel in the corrected image. The calculation formula is as follows: ； In the above formula, The coordinate position is The pixels in the temperature range The correction value, , The coordinates of the infrared image to be corrected are: The pixel response value, The coordinate position is The pixels in the temperature range The correction value, , The coordinate position is The pixels in the temperature range The weight value, , The coordinate position is The pixels in the temperature range The weight value, .

Images