An aerial target infrared radiation brightness calculation method based on pixel-by-pixel calibration

Abstract

The application discloses a kind of based on the calibration of each pixel air target infrared radiation brightness calculation method, including infrared detector, wherein: the calculation method includes calibration link, air target data acquisition link, air target radiation brightness calculation link;The radiation brightness calculation link: is according to the response gain coefficient matrix and bias matrix obtained by calibration, and the atmospheric path radiation, sky background radiation and transmittance on the path between data acquisition link obtained target and infrared measurement, and by the air infrared image target obtained from calculation model constructs air target infrared radiation brightness calculation model, and obtains the radiation brightness of target by calculation model.This method carries out infrared radiation calibration to detector focal plane array each pixel, establishes response gain coefficient and bias matrix, improves air target infrared radiation brightness calculation model from the difference between pixels, solves the problem that traditional model causes large error due to non-uniformity, and improves the precision of calculation.

Description

Technical Field

[0001] This invention relates to a method for calculating the infrared radiation brightness of an aerial target based on pixel-by-pixel calibration. Background Technology

[0002] In infrared radiation characteristic measurement, when the target is close to the measurement system, it is mainly imaged as an extended target on the focal plane array of the infrared detector. Its geometry is clearly discernible, and such targets are also called area targets. The infrared characteristics of extended targets are generally described using radiance, which refers to the power emitted by a radiation source per unit projected area in a specific direction within a unit solid angle. It characterizes the spatial distribution of infrared radiation energy of the extended target. Conducting infrared radiation characteristic measurements of aerial targets first requires infrared radiation calibration to establish a linear relationship between the output grayscale value of the infrared measurement system and the radiant amount of a standard infrared radiation source. Infrared radiation calibration determines the standard used in infrared radiation characteristic measurement, directly affecting the numerical accuracy of subsequent calculations of the target's true infrared characteristics from infrared image data. Currently, in engineering applications such as laboratory infrared radiation calibration and outdoor environmental infrared testing, most calibration methods use the averaging of grayscale values ​​across the entire image frame at different temperatures. This method is simple to operate, practical, and has strong anti-interference capabilities. He Yuanxing, Zhang Haoyuan, Si Wentao, and others proposed a general model for calculating the infrared radiance and intensity of extended-range and small targets in the air. This model assumes that the infrared radiance calibration response gain coefficient and bias of all pixels in the entire image spatial domain are equal and constant, and therefore uses background cancellation to eliminate the influence of ambient temperature changes on the bias. However, in practical work, we found that the infrared calibration gain coefficient and bias of different pixels in the image spatial domain are different, and as the infrared detector is used for a longer period of time, the difference in calibration values ​​increases continuously, and its influence can no longer be ignored. Simply using the average calibration result to calculate the infrared radiance of the target results in a very large error, which is difficult to meet the requirements of high-precision measurement. Summary of the Invention

[0003] The purpose of this invention is to propose a method for calculating the infrared radiation brightness of aerial targets based on pixel-by-pixel calibration. This method solves the problem that traditional methods for calculating the infrared radiation brightness of aerial targets neglect the non-uniformity of the infrared calibration response gain coefficient and bias of different pixels in the image spatial domain, resulting in excessive calculation errors and failing to meet the requirements of high-precision measurement.

[0004] To achieve the above objectives, the technical solution of the present invention is as follows:

[0005] A method for calculating the infrared radiance of an aerial target based on pixel-by-pixel calibration, used for measuring infrared radiation characteristics, includes an infrared detector, wherein: the calculation method includes a calibration step, an aerial target data acquisition step, and an aerial target radiance calculation step;

[0006] The calibration process involves dividing a standard blackbody into different temperature points to determine the linear relationship between the pixel-by-pixel response grayscale value of the infrared image and the infrared radiance of the target at different temperatures, thereby obtaining the response gain coefficient matrix and the bias matrix.

[0007] The aerial target data acquisition stage involves: stably tracking the aerial target to obtain an infrared image including the target and a clean sky background; simultaneously, using a solar radiometer, a scattering lidar, and atmospheric measurement equipment from a ground-based integrated meteorological station to collect atmospheric parameters in real time, such as atmospheric path radiation, sky background radiation, and transmittance along the path between the target and the infrared measurement.

[0008] The radiance calculation step involves constructing an infrared radiance calculation model for the aerial target based on the calibrated response gain coefficient matrix and bias matrix, as well as the atmospheric path radiation, sky background radiation, and transmittance along the path between the target and the infrared measurement obtained from the data acquisition step. The radiance of the target is then obtained from the calculation model.

[0009] The solution further includes the following steps for obtaining the response gain coefficient matrix and the bias matrix:

[0010] Step 1: Using the close-range extended source calibration method, place the standard blackbody surface source in front of the infrared detector so that the radiation source completely covers the entrance pupil of the infrared detector imaging system, acquire the image of the standard blackbody surface source, and adjust the blackbody to n different temperature points T. i Record different temperature points T i Image H ij ;

[0011] Where: n is not less than 2, i is an integer and 1≤i≤n, and j is the number of frames of the recorded image;

[0012] Step 2: Calculate the standard blackbody at n different temperature points T using Planck's formulas (1), (2), and (3). i Radiance value L i :

[0013]

[0014]

[0015]

[0016] in:

[0017] L i (T i f2, f1) represents a standard blackbody with a spectral frequency range of f2:f1 at temperature T.i The radiance value, in W / m² 2 ·sr;

[0018] h is Planck's constant, with a value of 6.6260693 × 10⁻⁶. -34 J.s;

[0019] k is the Boltzmann constant, with a value of 1.3806505 × 10⁻⁶. -23 J / K, where c is the speed of light, equals 2.99792458 × 10⁻⁶. 8 m / s;

[0020] T i Temperature, in degrees Fahrenheit;

[0021] f is the spectral frequency;

[0022] λ1:λ2 represents the range of infrared light response wavelengths;

[0023] Step 3: Apply formula (4) to the same temperature T i Calculate the average pixel grayscale value of images with different frame numbers.

[0024]

[0025] The linear relationship between the pixel-by-pixel response grayscale value of the infrared image and the infrared radiance of the target at different temperatures was calculated using the least squares method. The initial response gain coefficient matrix G0 and the initial bias matrix B0 obtained by formula (5) were then established.

[0026]

[0027] G0 and B0 are smoothed by mean filtering with a template size of 10×10 to obtain the final response gain coefficient matrix G and bias matrix B.

[0028] in:

[0029] x and y represent the image H respectively. i Pixel position coordinates along the x and y directions;

[0030] H ij (x,y) represents the grayscale values ​​of image pixels with different frame numbers at the same temperature;

[0031] H i (x,y) represents the average gray value of the pixel at (x,y) for different frame numbers of images at the same temperature;

[0032] G0(x,y) and B0(x,y) are the response gain coefficient and bias of the pixel at (x,y), respectively.

[0033] The solution further states that the atmospheric path radiation, sky background radiation, and transmittance along the path between the target and the infrared measurement are calculated using the MODTRAN radiative transfer model.

[0034] The solution further includes: the method for constructing the infrared radiation brightness calculation model for aerial targets is:

[0035] First, the area where the infrared image target is located is divided into region A, region B, and region C. Region A includes all the extended target and part of the sky background, with a total of N1 pixels. Region B is the outer extension of region A, generally extending outward by 2 pixels, with a total of N2 pixels. Among them, the rectangular ring region C, which is inside region B and outside region A, is the sky background, with a total of N2-N1 pixels; defined by formula (6). The average gray value of the sky background region C in the image is obtained by formulas (7) and (8). and

[0036]

[0037]

[0038] in:

[0039] and These are the calibration average response gain coefficient and average bias for the sky background region C, respectively;

[0040] H(x,y) is the gray value of the image at (x,y);

[0041] Second, construct a pixel-by-pixel calibration model for calculating the infrared radiance of an aerial target, as shown in formula (9), and calculate and invert the radiance L of the target. target :

[0042]

[0043] in:

[0044] G(x,y) and B(x,y) are the values ​​of the response gain coefficient matrix G and the bias matrix B at (x,y);

[0045] H target (x,y) represents the gray value of the target at (x,y);

[0046] L target (x,y) represents the radiant brightness of the airborne target calculated at (x,y);

[0047] L path Atmospheric path radiation;

[0048] L back Background radiation from the sky;

[0049] τ atm Transmittance.

[0050] The scheme further includes: the near-range extended source calibration method is as follows: a measurement system consisting of a cooled mid-wave infrared camera is used. The system has 640×512 pixels and a wavelength of 3.7-4.8um. ​​The radiance of a standard surface source blackbody located 420m away from the ground measurement system is measured and verified. The blackbody has a size of 600mm×600mm and an emissivity of 0.99.

[0051] The main advantages of the present invention are as follows: the method performs infrared radiation calibration on each pixel of the detector focal plane array, establishes the response gain coefficient and bias matrix, improves the calculation model of infrared radiation brightness of airborne targets by starting from the differences between pixels, solves the problem of large error caused by non-uniform calibration of traditional models, and improves the accuracy of the calculation method.

[0052] By calibrating the infrared radiation of each pixel in the infrared measurement system, the linear relationship between the grayscale response of a single pixel and the standard infrared radiation source is obtained. This improves the current method of calibrating the average value of the frame, which is commonly used in infrared characteristic testing. It also provides a calculation model for the infrared radiation brightness of aerial targets based on pixel-by-pixel calibration, which effectively overcomes the problem of large errors caused by the non-uniformity of calibration in traditional models. This improves the accuracy of the calculation method and has the characteristics of strong operability and wide applicability. It can be used as a general method for calculating the infrared radiation brightness of aerial targets.

[0053] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. Attached Figure Description

[0054] Figure 1 This is a flowchart of the process steps of the present invention;

[0055] Figure 2 A schematic diagram of a near-range extended source calibration method;

[0056] Figure 3 A map is created to delineate the region where the target is located in the infrared image;

[0057] Figure 4 Numerical plot of the response gain coefficient matrix G for Example 1;

[0058] Figure 5 Numerical plot of bias matrix B in Example 1;

[0059] Figure 6 This is a schematic diagram of the image processing method in Example 1. Detailed Implementation

[0060] A method for calculating the infrared radiance of an aerial target based on pixel-by-pixel calibration, used for measuring infrared radiation characteristics, includes an infrared detector, wherein: the calculation method includes a calibration step, an aerial target data acquisition step, and an aerial target radiance calculation step;

[0061] The calibration process involves dividing a standard blackbody into different temperature points to determine the linear relationship between the pixel-by-pixel response grayscale value of the infrared image and the infrared radiance of the target at different temperatures, thereby obtaining the response gain coefficient matrix and the bias matrix.

[0062] The aerial target data acquisition stage involves: stably tracking the aerial target to obtain an infrared image including the target and a clean sky background; simultaneously, using a solar radiometer, Mie scattering lidar, and ground-based integrated meteorological station atmospheric measurement equipment to collect atmospheric parameters in real time, obtaining atmospheric path radiation, sky background radiation, and transmittance along the path between the target and the infrared measurement.

[0063] The radiance calculation step involves constructing an infrared radiance calculation model for the aerial target based on the calibrated response gain coefficient matrix and bias matrix, as well as the atmospheric path radiation, sky background radiation, and transmittance along the path between the target and the infrared measurement obtained from the data acquisition step. The radiance of the target is then obtained from the calculation model.

[0064] See Figure 1 The steps for obtaining the response gain coefficient matrix and the bias matrix include:

[0065] Step 1: Using the close-range extended source calibration method, place the standard blackbody surface source in front of the infrared detector so that the radiation source completely covers the entrance pupil of the infrared detector imaging system, acquire the image of the standard blackbody surface source, and adjust the blackbody to n different temperature points T. i Record different temperature points T i Image H ij ;

[0066] Where: n is not less than 2, i is an integer and 1≤i≤n, and j is the number of frames of the recorded image;

[0067] Step 2: Calculate the standard blackbody at n different temperature points T using Planck's formulas (1), (2), and (3). i Radiance value L i :

[0068]

[0069]

[0070]

[0071] in:

[0072] L i (T i f2, f1) represents a standard blackbody with a spectral frequency range of f2:f1 at temperature T. i The radiance value (hereinafter abbreviated as L) i (Unit: W / m) 2 ·sr;

[0073] h is Planck's constant, with a value of 6.6260693 × 10⁻⁶. -34 J.s;

[0074] k is the Boltzmann constant, with a value of 1.3806505 × 10⁻⁶. -23 J / K, where c is the speed of light, equals 2.99792458 × 10⁻⁶. 8 m / s;

[0075] T i Temperature, in degrees Fahrenheit;

[0076] f is the spectral frequency;

[0077] λ1:λ2 represents the range of infrared light response wavelengths;

[0078] Step 3: Apply formula (4) to the same temperature T i Image per-pixel grayscale value H at different frame numbers ij Find the average value of (x, y).

[0079]

[0080] The linear relationship between the pixel-by-pixel response grayscale value of the infrared image and the infrared radiance of the target at different temperatures was calculated using the least squares method. The initial response gain coefficient matrix G0 and the initial bias matrix B0 obtained by formula (5) were then established.

[0081]

[0082] G0 and B0 are smoothed by mean filtering with a template size of 10×10 to obtain the final response gain coefficient matrix G and bias matrix B.

[0083] in:

[0084] x and y represent the image H respectively. i Pixel position coordinates along the x and y directions;

[0085] H ij (x,y) represents the grayscale values ​​of image pixels with different frame numbers at the same temperature;

[0086] H i (x,y) represents the average gray value of the pixel at (x,y) for different frame numbers of images at the same temperature;

[0087] G0(x,y) and B0(x,y) are the response gain coefficient and bias of the pixel at (x,y), respectively.

[0088] In the embodiment: the model for calculating the infrared radiation brightness of an aerial target is:

[0089] First, such as Figure 3 As shown, the area where the infrared image target is located is divided into an inner frame region A, an outer frame region B, and a rectangular ring region C between the inner and outer frames. Region A includes all the extended target and part of the sky background, with a total of N1 pixels. Region B is the outer extension of region A, generally extending outward by 2 pixels, with a total of N2 pixels. Among them, the rectangular ring region C, which is inside region B and outside region A, is the sky background, with a total of N2-N1 pixels; defined by formula (6). The average gray value of the sky background region C in the image is obtained by formulas (7) and (8). and

[0090]

[0091]

[0092]

[0093] in:

[0094] and These are the calibration average response gain coefficient and average bias for the sky background region C, respectively;

[0095] H(x,y) is the gray value of the image at (x,y);

[0096] Second, construct a pixel-by-pixel calibration model for calculating the infrared radiance of an aerial target, as shown in formula (9), and calculate and invert the radiance L of the target. target :

[0097]

[0098] in:

[0099] G(x,y) and B(x,y) are the values ​​of the response gain coefficient matrix G and the bias matrix B at (x,y);

[0100] H target (x,y) represents the gray value of the target at (x,y);

[0101] L target (x,y) represents the radiant brightness of the airborne target calculated at (x,y);

[0102] L path Atmospheric path radiation;

[0103] L back Background radiation from the sky;

[0104] τ atm Transmittance.

[0105] The atmospheric path radiation L along the path between the target and the infrared measurement is obtained. path Sky background radiation L back and transmittance τ atm It was calculated using the MODTRAN radiative transfer model, which is a known technique and will not be elaborated upon here.

[0106] The method of calibration using a near-range extended source involves using a measurement system consisting of a cooled mid-wave infrared camera with 640×512 pixels and a wavelength of 3.7-4.8µm. The system measures and verifies the radiance of a standard blackbody at a distance of 420m from the ground source measurement system. The blackbody has a size of 600mm×600mm and an emissivity of 0.99.

[0107] Here is a specific example:

[0108] Step 1: Place the standard blackbody directly in front of the measurement system, ensuring that the blackbody surface source covers the entrance pupil of the measurement system, such as... Figure 2 As shown, we consider the effect of atmospheric loss to be negligible at this point. Low-temperature calibration was performed, with the blackbody temperature successively adjusted to 50℃, 70℃, 90℃, and 110℃. The infrared measurement system recorded 10 frames of images for each temperature point, with a camera integration time of 500µm. The average of the image sets at the four temperature points was calculated using formula (4):

[0109] Step 2: Using Planck's formulas (2)-(4) for blackbody radiation, calculate the standard blackbody radiance at four temperature points: 50℃, 70℃, 90℃, and 110℃, which are 2.7685 W / m². 2 ·sr、5.0300W / m 2 ·sr、8.5704W / m 2 ·sr、13.8304W / m 2 Using the least squares method formula (5), the response gain coefficient matrix G and the bias matrix B are calculated, as follows: Figure 4 and Figure 5 As shown.

[0110] Step 3: After calibration, the blackbody was placed on the ground at a distance of 420m from the measurement system. The temperature of the blackbody was adjusted sequentially to 90℃, 110℃, 130℃, 150℃, and 200℃. The infrared measurement system used a staring image recording method, with a camera integration time of 500µm. During this process, atmospheric parameters were collected in real time using atmospheric measurement equipment such as a solar radiometer, Mie scattering lidar, and a ground-based integrated meteorological station. The atmospheric path radiation L at a distance of 420m was calculated using the MODTRAN model. path It is 0.2917 W / m 2 ·sr, transmittance τ atm It is 0.738.

[0111] Step 4: To verify the accuracy of the method and eliminate measurement errors of atmospheric equipment on background radiation, we used a blackbody image and blackbody radiance values ​​as a standard background for data verification. Figure 6 As shown, a 30-pixel × 30-pixel region is selected at (x, y) of the blackbody image at temperature T1 as the target to be measured, with (x, y) being the coordinates of the center point. The blackbody radiance value L1 is the theoretical radiance L of the target. target In a blackbody image at temperature T2, a rectangular region centered at (x,y) between 30 pixels × 30 pixels and 34 pixels × 34 pixels is taken as the standard background. The blackbody radiance value L2 is the standard background radiance L. back Based on the verification results compared with the traditional method shown in Table 1:

[0112] Table 1

[0113]

[0114] Blackbody images at temperatures T1 and T2 are used as different combinations of target and background. The average gray value of the background image is calculated from the images. Background calibration of average response gain coefficient and background average bias The radiance calculation model mentioned in this invention was used for verification calculations. The error and standard deviation data show that the method proposed in this invention has a significant improvement in calculation accuracy compared to traditional methods.

[0115] The above detailed description further illustrates the purpose, technical solution, implementation process, and beneficial effects of the invention. It should be understood that the above description is merely one application example and is not intended to limit the scope of protection of this invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for calculating the infrared radiance of an aerial target based on pixel-by-pixel calibration, used for measuring infrared radiation characteristics, comprising an infrared detector, characterized in that, The calculation method includes a calibration step, an aerial target data acquisition step, and an aerial target radiance calculation step. The calibration process involves dividing a standard blackbody into different temperature points to determine the linear relationship between the pixel-by-pixel response grayscale value of the infrared image and the infrared radiance of the target at different temperatures, thereby obtaining the response gain coefficient matrix and the bias matrix. The aerial target data acquisition stage involves: stably tracking the aerial target to obtain an infrared image including the target and a clean sky background; simultaneously, using a solar radiometer, a scattering lidar, and atmospheric measurement equipment from a ground-based integrated meteorological station to collect atmospheric parameters in real time, such as atmospheric path radiation, sky background radiation, and transmittance along the path between the target and the infrared measurement. The radiance calculation step involves constructing an infrared radiance calculation model for the aerial target based on the calibrated response gain coefficient matrix and bias matrix, as well as the atmospheric path radiation, sky background radiation, and transmittance along the path between the target and the infrared measurement obtained in the data acquisition step. The radiance of the target is then obtained from the calculation model. The model for calculating the infrared radiation of aerial targets is: First, the area containing the target in the infrared image is divided into region A, region B, and region C. Region A includes all extended targets and part of the sky background, totaling [number missing]. Region A consists of 100 pixels; Region B is an extension of Region A, typically extending outward by 2 pixels, and has a total of 100 pixels. There are [number] pixels; among which, the rectangular ring region C, located within region B and outside region A, is the sky background, totaling [number] pixels. One cell; defined by formula (6) The average gray value of the sky background region C in the image is obtained by formulas (7) and (8). and : （6） （7） （8） in: and These are the calibration average response gain coefficient and average bias for the sky background region C, respectively; For the image in The grayscale value at that location; Second, construct a pixel-by-pixel calibration model for calculating the infrared radiance of an aerial target, as shown in formula (9), and calculate and invert the radiance of the target. : （9） in: and For the response gain coefficient matrix and bias matrix exist The value at that location; For the target to be tested in The grayscale value at that location; In order to be in The calculated radiative brightness value of the airborne target; Atmospheric path radiation; Background radiation from the sky; Transmittance.

2. The method for calculating the infrared radiation brightness of an aerial target according to claim 1, characterized in that, The steps for obtaining the response gain coefficient matrix and the bias matrix include: Step 1: Using the close-range extended source calibration method, place the standard blackbody surface source in front of the infrared detector so that the radiation source completely covers the entrance pupil of the infrared detector imaging system, acquire the image of the standard blackbody surface source, and adjust the blackbody to n different temperature points. Record different temperature points Image ; Where: n is not less than 2, and i is an integer and j is the number of frames in the recorded image; Step 2: Calculate the standard blackbody at n different temperature points using Planck's formulas (1), (2), and (3). Radiance value : （1） （2） （3） in: express A standard blackbody in the spectral frequency range at temperature The radiance value, in units ; h is Planck's constant, and its value is... ; k is the Boltzmann constant, and its value is... c is the speed of light, equal to ; Temperature, in Kelvin; f is the spectral frequency; The range of infrared light response wavelengths; Step 3: Apply formula (4) to the same temperature Calculate the average pixel-by-pixel grayscale value of images with different frame numbers: （4） Where: j is the number of frames in the recorded image. max To record the maximum number of image frames; The linear relationship between the pixel-by-pixel response grayscale value of the infrared image and the infrared radiance of the target at different temperatures was calculated using the least squares method, and the initial response gain coefficient matrix obtained by formula (5) was established. and the initial bias matrix : （5） right and Mean filtering and smoothing are performed using a 10×10 template to obtain the final response gain coefficient matrix. and bias matrix ; in: x and y represent the image respectively. Pixel position coordinates along the x and y directions; These represent the grayscale values ​​of image pixels with different frame numbers at the same temperature. For images with different frame numbers at the same temperature The average gray level of the pixel; and They are respectively in The response gain coefficient and bias of the pixel; L1, L2...Ln are Li(Ti, f2, f1) in Formula 1: the radiance values ​​of a standard surface source blackbody at the 1st, 2nd, ..., nth different temperature points.

3. The method for calculating the infrared radiation brightness of an aerial target according to claim 1, characterized in that, The atmospheric path radiation, sky background radiation, and transmittance along the path between the target and the infrared measurement were calculated using the MODTRAN radiative transfer model.

4. The method for calculating the infrared radiation brightness of an aerial target according to claim 2, characterized in that, The calibration method using a near-range extended source is as follows: A measurement system consisting of a cooled mid-wave infrared camera is used. The system has 640×512 pixels and a wavelength of 3.7-4.8µm. The radiance of a standard blackbody surface source located 420m away from the ground measurement system is measured and verified. The blackbody has a size of 600mm×600mm and an emissivity of 0.99.

Images