Data Processing and Calibration Method for Circular Gradient Filter Spectroradiometer

Abstract

The present invention relates to a method for processing and calibrating data of a circular variable filter spectroradiometer, including: receiving detector data sent in packets by the instrument end, splicing them to obtain complete detector data; correcting the collected original data to obtain detector data after drift correction; performing wavelength registration and then passing registration verification to obtain precisely registered spectral data; comparing the precisely registered spectral data with spectra at different temperature points recorded within the interval, and obtaining the responsivity function of the target to be measured through linear interpolation. The present invention proposes a drift correction method based on covariance cross-correlation and cyclic offset correction, and on the basis of CVF spectral calibration, uses the gas absorption method to re-register the wavelength after drift correction, solving the problem of spectral distortion caused by drift during the rotation of the CVF; the present invention can solve the non-linear response problem brought about by a large measurement temperature range and a wide working wavelength band, ensuring the accuracy of the measurement results of the instrument.

Description

Technical Field

[0001] The present invention relates to the technical field of spectroradiometer measurement, and in particular to a data processing and calibration method for a circular variable filter spectroradiometer. Background Technique

[0002] A spectroradiometer is used to measure the spectral radiation characteristics of a radiation source and can characterize and distinguish optical signals by its spectral resolution ability. It is widely used in industries, scientific research, national defense and other fields, such as evaluating the emissivity of materials, the radiation characteristics of missile flames, and the field of atmospheric remote sensing. Due to its wide application in different fields, the spectroradiometer has become an indispensable measuring instrument.

[0003] Currently, spectroradiometers can be mainly classified into interference type, dispersion type, filter type, etc. according to the spectral resolution type. Each type of radiometer has its own unique characteristics and is therefore suitable for different application scenarios. Among them, interference-type and dispersion-type spectroradiometers are the two types that have developed most rapidly at home and abroad, and at the same time, the data processing research on these two types of radiometers is relatively mature.

[0004] The transmission wavelength of a circular variable filter (CVF) is linearly related to the angle. The spectroradiometer composed of it has advantages such as a wide spectral band and a large target temperature range. Therefore, the application scenarios of this type of spectroradiometer are also more extensive. Since different types of radiometers have different characteristics, their processing steps and methods are also different. Therefore, the data processing and calibration of a spectroradiometer usually need to be carried out based on the characteristics of the instrument itself. Adopting an effective data processing and calibration method for a CVF spectroradiometer is a prerequisite for ensuring the accuracy of measurement results.

[0005] Currently, the radiation calibration methods of spectroradiometers mainly include single-point method, two-point method and multi-point method, etc. The single-point method is suitable for the situation where the instrument resolution is low and the number of data points is small. The two-point method is suitable for the situation where the spectroradiometer has good linearity and the number of measured data points is large. Due to the characteristics of wide working wavelength band and wide target temperature range in the CVF type spectroradiometer, it brings a certain degree of non-linearity problem, making it impossible for traditional calibration methods to achieve accurate radiation calibration. Summary of the Invention

[0006] In order to solve the problem that the traditional method cannot be calibrated in a wide spectral band and a large target temperature range, the purpose of the present invention is to provide a data processing and calibration method for a circular variable filter spectroradiometer that can achieve accurate radiation calibration and ensure the accuracy of the instrument measurement results.

[0007] To achieve the above object, the present invention adopts the following technical solution: A data processing and calibration method for a circular variable filter spectroradiometer, the method includes the following steps in sequence:

[0008] (1) Data splicing: Receive the detector data sent in packets by the instrument end, splice it through data instruction parsing and verification, and turn it into complete detector data;

[0009] (2) Drift correction: Calculate the drift degree of the complete detector data through the covariance cross-correlation method, and correct the collected original data through cyclic offset correction and the calculated offset value to obtain the detector data after drift correction;

[0010] (3) Wavelength re-registration: Divided into two parts: wavelength registration and registration verification. After wavelength registration of the detector data after drift correction, and then through registration verification, obtain the accurately registered spectral data;

[0011] (4) Radiometric calibration: When performing infrared spectrum measurement, divide the temperature range of the measured target into n sub-ranges. Within the temperature range, measure and record the accurately registered spectral data S C i (λ) corresponding to n + 1 different blackbody temperatures. Compare the accurately registered spectral data S C i (λ) with the spectral data at different temperature points recorded within the range to determine the upper and lower limits of the temperature sub-range to which the target to be measured belongs; According to the response function calculated for the sub-range, obtain the response function of the target to be measured through linear interpolation, and substitute the response function into the calibration formula to complete radiometric calibration.

[0012] The specific steps of step (1) include the following steps:

[0013] (1a) Parse the data packet according to the instruction format, and use the method of regular expression registration to perform validity verification on the start and end check bits of the data packet instruction;

[0014] (1b) Parse the frame number and packet number in the data packet, ensure the continuity of the data through the frame number and packet number, compare the frame number and packet number in the valid data packet with the historically saved frame number and packet number, and trigger different processing mechanisms through different comparison results;

[0015] (1c) Parse the data in the valid data packet and update the size of the temporarily stored data volume;

[0016] (1d) Compare the size of the temporarily stored data volume with the data volume corresponding to the scanning rate to determine whether the detector data meets the data integrity requirements;

[0017] (1e) If the data integrity requirements are met, perform data splicing and output the complete detector data; otherwise, continue to receive the detector data packet.

[0018] The specific steps of step (2) are as follows:

[0019] (2a) After the rotation speed of the circular gradient filter reaches stability, perform multiple measurements, and use the average value of the multiple measurement data as the reference spectrum line;

[0020] (2b) Calculate the covariance between signals by the covariance cross-correlation method to determine the cross-correlation function, obtain the covariance cross-correlation function, determine the degree of drift through the covariance cross-correlation function, normalize the covariance cross-correlation function value, obtain the correlation degree matrix, and find the offset position with the highest correlation degree in the correlation degree matrix, which is the offset corresponding to the detector data. The calculation formula of the covariance cross-correlation function is as follows: The calculation formula is as follows:

[0021]

[0022] In the formula, y n is the reference signal, * represents the complex conjugate. In the formula, n and m respectively represent the relative displacements of the x and y signals, and N is the total offset of the two signals;

[0023] Normalize the covariance cross-correlation function:

[0024]

[0025] In the formula, is the normalized covariance cross-correlation function, is the unnormalized covariance cross-correlation function, are the covariance autocorrelation functions of the x and y signals respectively;

[0026] (2c) Through the calculated offset, use the cyclic offset method to finally correct the drift data to obtain the detector data after drift correction.

[0027] The specific steps of step (3) are as follows:

[0028] (3a) Use the starting position of the stepper motor saved historically to map the data sampling points of the detector data after drift correction to the step numbers of the stepper motor;

[0029] (3b) Use the relationship between the step number of the stepper motor and the transmission wavelength of the circular gradient filter to perform wavelength matching on the data to obtain the roughly registered data:

[0030] wavelength=k×motorindex+b

[0031] In the formula, wavelength is the wavelength value, motorindex is the step number of the stepper motor, and k and b are both calibration coefficients;

[0032] (3c) Calculate the theoretical CO transmittance curve using the MODTRAN radiative transfer model, and compare the CO absorption peaks in the coarsely registered data with the absorption peak positions in the theoretical CO transmittance to determine the degree of offset; 2 The CO absorption peaks in the coarsely registered data are 2 compared with the theoretical CO 2 absorption peaks in the transmittance to judge the degree of offset;

[0033] (3d) Calculate the coarsely registered spectral data based on the degree of offset; if there is an offset in the comparison result, calculate the actual starting value of the stepper motor through the offset amount, iteratively update the actual value, re - perform wavelength registration through the new value, and remove the invalid values in the interval of the re - registered data to obtain the finely registered spectral data.

[0034] Step (4) specifically includes the following steps:

[0035] (4a) During measurement, compare the measured target spectrum S m (λ) with the finely registered spectral data S C i (λ) corresponding to different blackbody temperatures within the temperature range to obtain the upper bound and the lower bound

[0036] of the sub - interval of the temperature to be measured. (4b) Integrate the spectral signals at the upper and lower bounds of the sub - interval of the temperature to be measured

[0037]

[0038] in its measured wavelength range, and use linear ratio to solve the linear interpolation coefficient α of the responsivity function. The calculation formula of the linear interpolation coefficient α is as follows: m where I is the integral of the measured spectral signal value within the wavelength response range of its detector;

[0039] (4c) Through linear interpolation calculation on the corresponding temperature sub - intervals K hot (λ), K cold (λ) found for the target to be measured, solve the responsivity function K m (λ) of the target to be measured. The linear interpolation formula is as follows:

[0040] K m (λ)=(1 - α)K cold (λ)+αK hot (λ)

[0041] where the responsivity functions K hot (λ), Kcold (λ) is calculated by measuring the spectral measurement data of a standard blackbody at the same temperature as the target to be measured. The calculation formula for the responsivity function is as follows:

[0042]

[0043] In the formula, ε BB is the emissivity of the target blackbody; L(λ, T BB ) is the radiance at wavelength λ and temperature T of the target blackbody BB ; L(λ, T Amb ) is the radiance at wavelength λ and temperature T of the internal environment of the instrument Amb ; L(λ, T IBB ) is the radiance at wavelength λ and temperature T of the internal blackbody IBB ; K(λ) is the responsivity function, S BB (λ) is the original signal for measuring the target blackbody; the radiance of the blackbody is calculated using Planck's formula:

[0044]

[0045] In the formula, the first radiation constant c 1 = 2πhc 2 , the second radiation constant c 2 = hc / k, λ is the wavelength, and T is the temperature;

[0046] (4d) Substitute the measured spectrum and the calculated responsivity function K m (λ) into the calibration formula to complete the radiometric calibration. The calibration formula is as follows:

[0047]

[0048] In the formula, S(λ) is the measured original signal, W(λ) is the radiance of the observed target related to wavelength λ; τ(λ, l) represents the atmospheric transmittance when the distance between the target and the radiometer is l; L(λ, T air ) is the radiance at environmental temperature T air , and L(λ, T 0 ) is the radiance at the internal reference blackbody temperature T 0 .

[0049] In step (1b), the different processing mechanisms include:

[0050] (1b1) When the frame numbers in the data packet are the same and the packet numbers are consecutive, it is determined that the incoming data packet is a continuous data packet. The data of this data packet will be temporarily saved, and the historically saved packet number will be incremented by 1;

[0051] (1b2) When the frame numbers in the data packet are different and the packet number is 0, it is determined that the received data is a completely new frame of data. The temporarily saved data will be cleared, and the frame number and packet number will be updated;

[0052] (1b3) When the frame number and packet number in the data packet are determined to be other cases than those in step (1b1) and step (1b2), it is determined that the data packet is a discontinuous invalid data packet. The data packet will be discarded, the temporarily saved data will also be cleared, and the historically saved frame number and packet number will be updated, waiting for a new frame of data packet.

[0053] As can be seen from the above technical solutions, the beneficial effects of the present invention are as follows: First, the present invention proposes a drift correction method based on covariance cross-correlation and cyclic offset correction, and uses the gas absorption method to re-register the wavelengths after drift correction on the basis of CVF spectral calibration, solving the problem of spectral distortion caused by drift during the rotation of CVF; Second, the present invention proposes a responsivity radiometric calibration method based on partition linearity, which can solve the non-linear response problem brought about by the large measurement temperature range and wide working wavelength band that cannot be solved by traditional methods, ensuring the accuracy of the measurement results of the instrument. Brief Description of the Drawings

[0054] Figure 1 is the flowchart of the method of the present invention;

[0055] Figure 2 is the schematic diagram before drift correction after data splicing in Embodiment 1 of the present invention;

[0056] Figure 3 is the schematic diagram after drift correction after data splicing in Embodiment 1 of the present invention;

[0057] Figure 4 is the comparison diagram of the absorption peak in the spectrum after wavelength re-registration in Embodiment 1 of the present invention and the CO 2 transmittance calculated by MODTRAN;

[0058] Figure 5 is the comparison diagram of the theoretical value and the actual value after radiometric calibration of the 600 °C blackbody in Embodiment 1 of the present invention. Detailed Embodiment

[0059] As Figure 1 shown, a method for data processing and calibration of a circular variable filter spectro-radiometer, the method includes the following steps in sequence:

[0060] (1) Data splicing: Receive the detector data sent by the instrument end in packets, splice it through data instruction parsing and verification, and turn it into complete detector data;

[0061] (2) Drift correction: Calculate the drift degree of the complete detector data through the covariance cross-correlation method. Correct the acquired raw data through cyclic offset correction and the calculated offset value to obtain the detector data after drift correction;

[0062] (3) Wavelength re-registration: It is divided into two parts: wavelength registration and registration verification. After wavelength registration of the detector data after drift correction and then through registration verification, obtain the precisely registered spectral data;

[0063] (4) Radiometric calibration: When performing infrared spectral measurement, divide the temperature range of the measured target into n sub-ranges. Within the temperature range, measure and record the precisely registered spectral data S C i (λ) corresponding to n + 1 different blackbody temperatures. Compare the precisely registered spectral data S C i (λ) with the spectral data at different temperature points recorded within the range to determine the upper and lower limits of the temperature sub-range to which the target to be measured belongs; According to the response function calculated for the sub-range, obtain the response function of the target to be measured through linear interpolation, and substitute the response function into the calibration formula to complete the radiometric calibration.

[0064] The specific steps of step (1) include the following steps:

[0065] (1a) Analyze the data packet according to the instruction format, and use the method of regular expression registration to perform validity verification on the start and end check bits of the data packet instruction;

[0066] (1b) Analyze the frame number and packet number in the data packet, ensure the continuity of the data through the frame number and packet number, compare the frame number and packet number in the valid data packet with the frame number and packet number saved historically, and trigger different processing mechanisms through different comparison results;

[0067] (1c) Analyze the data in the valid data packet and update the size of the temporarily stored data volume;

[0068] (1d) Compare the size of the temporarily stored data volume with the data volume corresponding to the scanning rate to determine whether the detector data meets the data integrity requirement;

[0069] (1e) If the data integrity requirement is met, perform data splicing and output the complete detector data; otherwise, continue to receive the detector data packet.

[0070] In step (1b), the different processing mechanisms include:

[0071] (1b1) When the frame numbers in the data packet are the same and the packet numbers are consecutive, it is determined that the incoming data is a consecutive data packet. The data of this data packet will be temporarily saved, and the historically saved packet number will be incremented by 1;

[0072] (1b2) When the frame numbers in the data packet are different and the packet number is 0, it is determined that the incoming data is a completely new frame of data. The temporarily saved data will be cleared, and the frame number and packet number will be updated;

[0073] (1b3) When the frame number and packet number in the data packet are determined to be other cases than those in step (1b1) and step (1b2), it is determined that the data packet is a non-consecutive invalid data packet. This data packet will be discarded, the temporarily saved data will also be cleared, and the historically saved frame number and packet number will be updated, waiting for a new frame of data packet.

[0074] The specific steps of step (2) are as follows:

[0075] (2a) After the rotation speed of the circular gradient filter reaches stability, multiple measurements are taken, and the average value of the multiple measurement data is used as the reference spectrum line;

[0076] (2b) The covariance between signals is calculated by the covariance cross-correlation method to determine the cross-correlation function, obtaining the covariance cross-correlation function. The degree of drift is determined through the covariance cross-correlation function, the covariance cross-correlation function values are normalized to obtain the correlation degree matrix, and the offset position with the highest correlation degree in the correlation degree matrix is found, which is the offset corresponding to the detector data. The formula for the covariance cross-correlation function is as follows:

[0077]

[0078] In the formula, y n is the reference signal, * represents the complex conjugate. In the formula, n and m respectively represent the relative displacements of the x and y signals, and N is the total offset of the two signals;

[0079] Normalize the covariance cross-correlation function:

[0080]

[0081] In the formula, is the normalized covariance cross-correlation function, is the unnormalized covariance cross-correlation function, are the covariance autocorrelation functions of the x and y signals respectively;

[0082] (2c) Through the calculated offset, the cyclic offset method is used to finally correct the drift data, obtaining the detector data after drift correction.

[0083] The step (3) specifically comprises the following steps:

[0084] (3a) Using the historically saved starting position of the stepper motor, the drift-corrected detector data is mapped between the data sampling point and the number of stepper motor steps;

[0085] (3b) Using the relationship between the number of stepper motor steps and the wavelength of the circular gradient filter, the wavelength of the data is matched to obtain the rough registration data:

[0086] wavelength=k×motorindex+b

[0087] In the formula, wavelength is the wavelength value, motorindex is the number of steps of the stepper motor, k, b 均 is the calibration coefficient;

[0088] (3c) Calculate the theoretical CO using the MODTRAN radiative transfer model 2 Transmittance curve, CO in the rough registration data 2 Absorption peak and theoretical CO 2 The absorption peak positions in the transmittance are compared to determine the degree of shift;

[0089] (3d) According to the degree of offset, the roughly aligned spectral data is calculated; if there is an offset in the comparison result, the actual starting value of the stepper motor is calculated by the offset, the actual value is iteratively updated, the wavelength is re-aligned with the new value, and the invalid values ​​in the interval area of ​​the re-aligned data are eliminated, so as to obtain the precisely aligned spectral data.

[0090] The step (4) specifically comprises the following steps:

[0091] (4a) During measurement, the target spectrum S m (λ) is precisely aligned with the spectral data S corresponding to different blackbody temperatures within the temperature range C i (λ) to obtain the upper bound of the target temperature sub-interval to be measured With the lower bound

[0092] (4b) Spectral signals of the upper and lower bounds of the temperature sub-interval to be measured Integrate within the measurement wavelength range and use the linear ratio to solve the linear interpolation coefficient α of the response function. The calculation formula of the linear interpolation coefficient α is as follows:

[0093]

[0094] In the formula, I m It is the integral of the measured spectral signal value within the wavelength response range of its detector; is the integral of the upper and lower bound temperature spectral signal values within the detector wavelength response range over the measurement sub-interval;

[0095] (4c) Find the corresponding temperature sub-interval K for the target to be measured hot (λ), K cold (λ) through linear interpolation calculation to solve for the responsivity function K m (λ) of the target to be measured. The linear interpolation formula is as follows:

[0096] K m (λ) = (1 - α)K cold (λ) + αK hot (λ)

[0097] In the formula, the responsivity function K of the sub-interval hot (λ), K cold (λ) is calculated from the spectral measurement data of the standard blackbody at the same temperature as the target to be measured. The formula for calculating the responsivity function is as follows:

[0098]

[0099] In the formula, ε BB is the emissivity of the target blackbody; L(λ, T BB ) is the radiance at wavelength λ and temperature T of the target blackbody BB ; L(λ, T Amb ) is the radiance at wavelength λ and temperature T of the internal environment of the instrument Amb ; L(λ, T IBB ) is the radiance at wavelength λ and temperature T of the internal blackbody IBB ; K(λ) is the responsivity function, S BB (λ) is the original signal of the measured target blackbody; the radiance of the blackbody is calculated using Planck's formula:

[0100]

[0101] In the formula, the first radiation constant c 1 = 2πhc 2 , the second radiation constant c 2 = hc / k, λ is the wavelength, and T is the temperature;

[0102] (4d) Substitute the measured spectrum and the calculated responsivity function K m (λ) into the calibration formula to complete the radiometric calibration. The calibration formula is as follows:

[0103]

[0104] Wherein, S(λ) is the original measured signal, and W(λ) is the radiance of the observed target related to the wavelength λ; τ(λ, l) represents the atmospheric transmittance when the distance between the target and the radiometer is l; L(λ, T air ) is the radiance at the environmental temperature of T air , and L(λ, T 0 ) is the radiance at the internal reference blackbody temperature of T 0 .

[0105] Example 1

[0106] The spectral wavelength range of the CVF spectroradiometer is 1.3 - 14.3 μm, the spectral resolution is less than 2% of the wavelength, and InSb and MCT detectors are used to collect signals within its wavelength range. The CVF rotation speed is set to 1 Hz, and a narrow field of view (NFOV) with a field of view angle of 7.5 mrad is used to measure a 600 °C blackbody at a horizontal distance of 3 m. The measured signal obtained by splicing the data is compared with the reference signal, and the results before and after drift correction are respectively as shown in Figure 2 、 3 . Through the comparison of actual offset and covariance cross-correlation calculation, the correction error is 1 sampling point.

[0107] Using the above steps to continue wavelength re-registration of the corrected measured signal, the obtained spectral curve is compared with the absorption peak in the CO 2 transmittance calculated by MODTRAN, as shown in Figure 4 . After wavelength re-registration, the average error of the result has decreased from 0.019 μm to 0.004 μm.

[0108] The re-registered spectrum is radiometrically calibrated, and the calibration result is compared with the theoretical Planck function. The result is as shown in Figure 5 . Using the Planck theory formula and the least squares method, the equivalent temperature of the blackbody at different temperatures in the entire wavelength range is inversely deduced. The equivalent temperature is 197.4 °C, and the calibration accuracy is better than 98%.

[0109] In summary, the present invention proposes a drift correction method based on covariance cross-correlation and cyclic offset correction, and on the basis of CVF spectral calibration, uses the gas absorption method to re-register the wavelength after drift correction, solving the problem of spectral distortion caused by drift during the rotation of the CVF; the present invention proposes a responsivity radiometric calibration method based on piecewise linearity, which can solve the non-linear response problem brought about by a large measurement temperature range and a wide working wavelength band that cannot be solved by traditional methods, ensuring the accuracy of the measurement results of this instrument.

[0110] The foregoing has shown and described the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and what is described in the above embodiments and the specification is only the principle of the present invention. Without departing from the spirit and scope of the present invention, various changes and improvements will occur to the present invention, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection required by the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for processing and calibrating spectral radiometer data of a circular gradient filter, characterized in that: This method includes the following steps in sequence: (1) Data splicing: Receive the detector data sent in packets by the instrument end, splice it through data instruction parsing and verification, and turn it into complete detector data; (2) Drift correction: Calculate the drift degree of the complete detector data by the covariance cross-correlation method, and correct the collected original data through cyclic offset correction and the calculated offset value to obtain the detector data after drift correction; (3) Wavelength re-registration: It is divided into two parts: wavelength registration and registration verification. After wavelength registration of the detector data after drift correction, and then through registration verification, obtain the precisely registered spectral data; (4) Radiometric calibration: When performing infrared spectral measurement, the temperature range of the target to be measured is divided into n sub-ranges. Within the temperature range, the accurately registered spectral data S C i (λ) corresponding to n + 1 different blackbody temperatures are measured and recorded. The accurately registered spectral data S C i (λ) is compared with the spectra of different temperature points recorded within the range to determine the upper and lower limits of the temperature sub-range to which the target to be measured belongs; according to the responsivity function calculated for the sub-range, the responsivity function of the target to be measured is obtained by linear interpolation, and the responsivity function is substituted into the calibration formula to complete the radiometric calibration.

2. The method for processing and calibrating spectral radiometer data of a circular gradient filter according to claim 1, characterized in that: The specific steps of the step (1) include the following steps: (1a) Parse the data packet according to the instruction format, and use the method of regular expression registration to perform validity verification on the start and end check bits of the data packet instruction; (1b) Parse the frame number and packet number in the data packet, ensure the continuity of the data through the frame number and packet number, compare the frame number and packet number in the valid data packet with the historically saved frame number and packet number, and trigger different processing mechanisms through different comparison results; (1c) Parse the data in the valid data packet and update the size of the temporarily stored data volume; (1d) Compare the size of the temporarily stored data volume with the data volume corresponding to the scanning rate to determine whether the detector data meets the data integrity requirements; (1e) If the data integrity requirements are met, perform data splicing and output the complete detector data, otherwise, continue to receive the detector data packet.

3. The method for processing and calibrating spectral radiometer data of a circular gradient filter according to claim 1, characterized in that: The specific steps of the step (2) include the following steps: (2a) After the rotation speed of the circular gradient filter reaches stability, perform multiple measurements, and use the average value of the multiple measurement data as the reference spectral line; (2b) Calculate the covariance between signals through the covariance cross-correlation method to determine the cross-correlation function, obtain the covariance cross-correlation function, determine the degree of drift through the covariance cross-correlation function, normalize the covariance cross-correlation function values to obtain the correlation degree matrix, and find the offset position with the highest correlation degree in the correlation degree matrix, which is the offset corresponding to the detector data. The covariance cross-correlation function has the following calculation formula: where y n is the reference signal, * represents the complex conjugate. In the formula, n and m respectively represent the relative displacements of the two signals x and y, and N is the total offset of the two signals; Normalize the covariance cross-correlation function: In the formula, is the normalized covariance cross-correlation function, is the unnormalized covariance cross-correlation function, are the covariance autocorrelation functions of the x and y signals respectively; (2c) Through the calculated offset, use the cyclic offset method to perform the final correction on the drift data to obtain the detector data after drift correction.

4. The method for processing and calibrating spectral radiometer data of a circular gradient filter according to claim 1, characterized in that: The specific steps of the step (3) include the following steps: (3a) Use the historically saved starting position of the stepper motor to map the data sampling points of the detector data after drift correction to the stepper motor step numbers; (3b) Use the relationship between the stepper motor step number and the transmission wavelength of the circular gradient filter to perform wavelength matching on the data to obtain the roughly registered data: wavelength = k × motorindex + b In the formula, wavelength is the wavelength value, motorindex is the stepper motor step number, and k and b are both calibration coefficients; (3c) Calculate the theoretical CO transmittance curve using the MODTRAN radiative transfer model, and compare the CO absorption peaks in the coarsely registered data with the absorption peak positions in the theoretical CO transmittance to determine the degree of offset; 2 The CO 2 absorption peaks in the coarsely registered data are compared with the absorption peak positions in the theoretical CO 2 transmittance to determine the degree of offset; (3d) Calculate the roughly registered spectral data according to the offset degree; if there is an offset in the comparison result, calculate the actual starting value of the stepper motor through the offset amount, iteratively update the actual value, perform wavelength registration again through the new value, and remove the invalid values in the interval area of the re-registered data, so as to obtain the precisely registered spectral data.

5. The method for processing and calibrating circular variable filter spectroradiometer data according to claim 1, characterized in that: the specific steps of step (4) include the following steps: (4a) During measurement, the measured target spectrum S m (λ) is compared with the precisely registered spectral data S C i (λ) corresponding to different blackbody temperatures within the temperature range to obtain the upper bound of the sub-range of the target temperature to be measured and the lower bound (4b) Spectral signals at the upper and lower bounds of the sub-interval of the temperature to be measured Integrate within its measurement wavelength interval, and use the linear ratio to solve for the linear interpolation coefficient α of the responsivity function. The calculation formula for the linear interpolation coefficient α is as follows: where, I m is the integral of the measured spectral signal value within the wavelength response range of its detector; is the integral of the spectral signal values at the upper and lower bound temperatures of the measurement sub-range within the wavelength response range of its detector; (4c) The corresponding temperature sub-interval K found for the target to be measured hot (λ), K cold (λ) Perform linear interpolation calculation to solve and obtain the responsivity function K of the target to be measured m (λ), and the linear interpolation formula is as follows: K m L(λ)=(1 - α)K cold + αK hot L(λ) In the formula, the responsivity functions K hot (λ) and K cold (λ) are calculated from the spectral measurement data of a standard blackbody at the same temperature as the target to be measured. The calculation formula for the responsivity function is as follows: where ε BB is the emissivity of the target blackbody; L(λ, T BB ) is the radiance at wavelength λ and temperature T of the target blackbody BB ; L(λ, T Amb ) is the radiance at wavelength λ and temperature T of the internal environment of the instrument Amb ; L(λ, T IBB ) is the radiance at wavelength λ and temperature T of the internal blackbody IBB ; K(λ) is the responsivity function, and S BB (λ) is the original signal for measuring the target blackbody; the radiance of the blackbody is calculated using the Planck formula: wherein, the first radiation constant c 1 = 2πhc 2 , the second radiation constant c 2 = hc / k, λ is the wavelength, and T is the temperature; (4d) Substitute the measured spectrum and the calculated responsivity function K m (λ) into the calibration formula to complete the radiometric calibration. The calibration formula is as follows: wherein, S(λ) is the original measured signal, W(λ) is the radiance of the observed target related to the wavelength λ; τ(λ, l) represents the atmospheric transmittance when the distance between the target and the radiometer is l; L(λ, T air ) is the radiance at an ambient temperature of T air , and L(λ, T 0 ) is the radiance at the internal reference blackbody temperature T 0 .

6. The method for processing and calibrating circular variable filter spectroradiometer data according to claim 2, characterized in that: in step (1b), the different processing mechanisms include: (1b1) When the frame numbers in the data packet are the same and the packet numbers are consecutive, it is determined that the incoming data is a continuous data packet, the data of this data packet will be temporarily saved, and the historically saved packet number will be incremented by 1; (1b2) When the frame numbers in the data packet are different and the packet number is 0, it is determined that the incoming data is a brand-new frame of data, the temporarily saved data will be cleared, and the frame number and packet number will be updated; (1b3) When the frame number and packet number in the data packet are determined to be other situations except step (1b1) and step (1b2), it is determined that this data packet is a non-continuous invalid data packet, this data packet will be discarded, the temporarily saved data will also be cleared, and the historically saved frame number and packet number will be updated, waiting for a new frame of data packet.

Images