Noctilucent remote sensing satellite in-orbit relative radiometric calibration method and system based on radiation reference transfer

Abstract

The invention discloses a noctilucent remote sensing satellite in-orbit relative radiation calibration method and system based on radiation reference transfer, and the method comprises the steps: extracting a daytime radiation reference, and obtaining the daytime radiation reference of a noctilucent remote sensing satellite; constructing a radiation transfer model among different imaging parameters of the noctilucent remote sensing satellite, and obtaining the radiation transfer model and coefficient among the different imaging parameters of the noctilucent remote sensing satellite; transmitting the daytime radiation reference of the noctilucent remote sensing satellite to the night to obtain an in-orbit relative radiation correction model of the night image of the noctilucent remote sensing satellite, and completing the in-orbit relative radiation calibration of the noctilucent remote sensing satellite. The method is suitable for in-orbit relative radiometric calibration of area array type noctilucent remote sensing satellites, and has the advantages of being high in precision, more convenient and low in cost.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing earth observation, and provides an on-orbit relative radiometric calibration method for a night imaging remote sensing satellite. In particular, for the problem that the area array night light remote sensing satellite payload cannot obtain the night light calibration benchmark at night, an on-orbit relative radiometric calibration method and system for the night light remote sensing satellite based on radiation benchmark transfer are provided. Background Art

[0002] In recent years, with the rapid development of satellite, sensor and data processing technologies, the trend of diversified development of remote sensing satellites has become more obvious, and new types of remote sensing satellites such as video satellites and night light satellites have emerged. Among them, the night light remote sensing satellite has become one of the hotspots of recent development. The night light remote sensing satellite captures the weak radiation information on the earth's surface at night and obtains the remote sensing image of the earth's surface at night, which can more directly monitor human activities. It has become an important branch of remote sensing space information acquisition and has important research significance.

[0003] On-orbit radiometric calibration is the core link to ensure the radiation quality of satellite remote sensing products and the premise for the effective application of remote sensing images. The earth observation data obtained from remote sensing satellites can only obtain various quantitative radiation information of ground objects through radiometric calibration, and then the spectral reflectance characteristics and spectral radiation characteristics of ground objects can be inverted, so as to correctly interpret and effectively utilize satellite remote sensing images. Although the radiation imaging parameters of the remote sensing satellite were calibrated in the laboratory before launch, during the satellite launch process, stress release, as well as during the long-term operation of the satellite, factors such as the space thermal environment, device aging, and component movement will cause changes in the radiation imaging parameters, and on-orbit radiometric calibration is urgently needed.

[0004] On-orbit relative calibration requires a high-precision radiation benchmark that can cover all detector elements to calibrate the response model relationship between the detector elements of the sensor. According to the methods for obtaining different radiation benchmarks, different on-orbit relative radiometric calibration methods have been developed, such as on-board calibration, uniform field calibration, statistical calibration, and yaw calibration methods. However, the above methods cannot be directly applied to night light remote sensing satellites. The on-board calibration method depends on on-board calibration equipment and cannot be implemented for night light satellites without on-board calibration equipment. The uniform field calibration method depends on large-area uniform ground objects with different surface reflectivities and good uniformity, such as the ocean, desert, clouds, snow, etc. There is no uniform light field or natural light field that meets the above conditions at night, so the night light remote sensing satellite cannot be directly applied. The statistical calibration method is limited by the accumulation of a large number of image samples and a reasonable brightness distribution, and is only effective for swing-scan or line-array push-broom remote sensing payloads. The yaw radiometric calibration is applicable to line-array push-broom remote sensing payloads, and both are difficult to apply to area-array night light remote sensing payloads. Therefore, studying the on-orbit radiometric calibration technology of area-array night light remote sensing satellites is of great significance for improving the image quality of night light satellites and ensuring the efficient application of night light satellite data in various social economics. Summary of the Invention

[0005] The object of the present invention is to provide an on-orbit relative radiometric calibration method for a night-light remote sensing satellite based on radiation reference transfer, which solves the problem that the night-light remote sensing satellite cannot obtain a calibration reference at night by transferring the extracted daytime radiation reference to the night, and realizes the on-orbit relative radiometric calibration of the night-light remote sensing satellite.

[0006] The technical solution of the present invention provides an on-orbit relative radiometric calibration method for a night-light remote sensing satellite based on radiation reference transfer, including the following steps:

[0007] Step 1, obtain daytime calibration imaging data and extract the daytime radiation reference, where the daytime radiation reference is expressed as the relative radiation coefficient of each detector element of the night-light remote sensing satellite;

[0008] Step 2, construct a radiation transfer model between different imaging parameters of the night-light remote sensing satellite to realize the radiation transfer between images with any imaging parameters of the night-light remote sensing satellite;

[0009] Step 3, based on the radiation transfer model in Step 2, transfer the extracted daytime radiation reference of the night-light remote sensing satellite to the night to obtain the on-orbit relative radiation coefficient of the night-light remote sensing satellite;

[0010] Step 4, extract image samples with different imaging parameters from the laboratory radiometric calibration image data or on-orbit imaging data before the launch of the night-light remote sensing satellite, and solve the relationship coefficients of the radiation transfer model;

[0011] Step 5, substitute the daytime radiation coefficient of the night-light remote sensing satellite obtained in Step 1 and the relationship coefficients of the radiation transfer model obtained in Step 4 into the radiation transfer model in Step 3 to realize the relative radiometric correction of the night-time image data of the night-light remote sensing satellite and complete the on-orbit relative radiometric calibration of the night-light remote sensing satellite.

[0012] Further, in Step 1, according to the actual resolution of the night-light remote sensing satellite, large-area uniform ground objects with different reflectivities that meet the following requirements are selected globally to obtain daytime calibration imaging data; the large-area uniform ground objects should meet the following conditions:

[0013] Condition (1): The brightness of the large-area uniform ground objects with different reflectivities can cover at least 3 different brightness levels in the dynamic range of the night-light sensor;

[0014] Condition (2): The large-area uniform ground objects cover all the imaging detector elements of the night-light remote sensing satellite;

[0015] Condition (3): The non-uniformity of the uniform field data used for daytime calibration itself should be higher than the calibration accuracy, otherwise the requirements of the calibration accuracy cannot be met.

[0016] Further, the specific implementation of the extraction of the daytime radiation reference includes:

[0017] First, an on-orbit radiation calibration model of a night-light remote sensing satellite is constructed, wherein the on-orbit radiation calibration model of a night-light remote sensing satellite is a first-order linear model or a multi-order nonlinear model;

[0018] Then, for the i-th imaging detector of the night-light remote sensing satellite, based on the acquired daytime calibration image data, the least squares method is used to solve the on-orbit radiation calibration model of the night-light remote sensing satellite to obtain the relative radiation coefficient of the i-th imaging detector, i∈[0,M], M is the number of night-light remote sensing satellite detectors;

[0019] Finally, the relative radiation coefficients of all detectors of the night light remote sensing satellite are obtained, that is, the daytime radiation benchmark is extracted.

[0020] Furthermore, when the on-orbit radiation calibration model of the night light remote sensing satellite is a first-order linear model, its calculation formula is as follows:

[0021]

[0022] Where j is the number of large-area uniform fields with different reflectivity, DN i,j is the radiation quantization value of the i-th detector of the j-th uniform field night light remote sensing satellite, is the grayscale mean of all imaging detectors of the jth uniform field night light remote sensing satellite, a i and b i is the relative radiation coefficient of the i-th detector of the night light remote sensing satellite.

[0023] Furthermore, the specific implementation of step 2 is as follows:

[0024] Step 2.1: Based on step 1, establish the radiation calibration model of the night light remote sensing satellite under different imaging parameters: When the night light remote sensing satellite is imaged under different imaging parameters, its calibration model is:

[0025]

[0026] Where a H 、b H 、a L 、b L Corresponding to the radiation coefficient under different imaging parameters, DN H DN L is the image grayscale value of the night light remote sensing satellite under different imaging parameters, is the mean grayscale value of the image under different imaging parameters, and the subscripts H and L are different imaging parameters, including low gain, high gain, low number of steps, and high number of steps.

[0027] Step 2.2, define the radiation response model relationship between different imaging parameters of the night light remote sensing satellite: According to the imaging characteristics of the night light remote sensing satellite, there is a certain model relationship between the radiation responses of its different imaging parameters, and there is an unknown order polynomial model relationship between the image gray values.

[0028] DN H = B0 + B1×DN L + B2×DN L 2 +…+ B n ×DN L n (IV) Furthermore, the gray mean values of different imaging images also have:

[0029]

[0030] In the formula, n is the polynomial order, n is greater than or equal to 1; B0, B1, B2…B n are the model relationship coefficients;

[0031] Step 2.3, combine formulas (II)-(V), and then the model relationship between the gray values DN H 、DN L of the night light remote sensing satellite's different imaging parameters can be established, and the radiation transfer between the images of any imaging parameters of the night light remote sensing satellite can be realized.

[0032] Furthermore, the specific implementation method of Step 2.3 is as follows:

[0033] Eliminate a H 、b H through the polynomial relationship, establish the model relationship between DN H and , and the radiation transfer of images with different imaging parameters is realized through a L 、b L in the model; when n = 2, the conversion model of the night light remote sensing satellite's images under different imaging parameters at night is as follows:

[0034]

[0035] When n > 2, the relationship between DN H and can be solved by combining formulas (II)-(V).

[0036] Furthermore, determine the order n by any one of the following two methods:

[0037] (1) Based on the satellite design value;

[0038] (2) Based on the optimal fitting of the satellite imaging data, take the polynomial order when the sum of the squares of the fitting residuals is the smallest.

[0039] Further, the specific implementation of step 3 is as follows:

[0040] In step 1, the relative radiation coefficients a i and b i of each detector element of the night light remote sensing satellite under the daytime imaging parameters have been solved through the daytime radiometric calibration data, that is, the daytime radiation benchmark, where i ∈ [0, M] and M is the number of detector elements of the night light remote sensing satellite; based on the radiation transfer model between different imaging parameters of the night light remote sensing satellite established in step 2, the daytime radiation benchmark can be transferred to the night; let a L = a i , b L = b i , substitute into formula (six), that is, transfer the daytime radiation benchmark a i , b i to the night; when n = 2, the transfer model of the relative radiation coefficients a L , b L of each detector element of the night light remote sensing satellite to the night is as follows.

[0041]

[0042] When n > 2, solve by combining formulas (two) - (five).

[0043] Further, in step 4, the number of laboratory radiometric calibration image data is not less than 50 scenes, and the number of on-orbit imaging data is not less than 100 scenes.

[0044] The present invention also provides a night light remote sensing satellite on-orbit relative radiometric calibration system based on radiation benchmark transfer, including:

[0045] A processor and a memory, the memory is used to store program instructions, and the processor is used to call the stored instructions in the memory to execute the night light remote sensing satellite on-orbit relative radiometric calibration method based on radiation benchmark transfer as described in the above technical solution.

[0046] Compared with the prior art, the present invention has the following characteristics and beneficial effects:

[0047] (1) It does not require the night light remote sensing satellite to have on-board calibration processing capabilities.

[0048] (2) It avoids the need for a uniform light source on the ground at night and can calibrate the night light image through the daytime radiation benchmark.

[0049] (3) It does not require direct calibration of night data.

[0050] Based on the imaging characteristics of the night-light remote sensing satellite, by establishing a radiation reference transfer model from daytime to night-time for the night-light remote sensing satellite, and using the daytime calibration imaging data of the night-light remote sensing satellite, the calibration of night-time image data is achieved; it avoids the high demand of the night-light remote sensing satellite for a large-area uniform calibration reference at night, solves the problem of no calibration reference for the relative radiometric calibration of the night-light remote sensing satellite in orbit, and improves the quality of the night-light satellite images. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a schematic diagram of the linear radiometric calibration model and fitting residuals of the 2,554,586th detector element of the night-light remote sensing satellite in the embodiment of the present invention;

[0052] Figure 2 It is a schematic diagram of the radiation transfer model (second-order polynomial model) between the low-high gain images of the night-light remote sensing satellite in the embodiment of the present invention.

[0053] Figure 3 It is a comparison chart of the effects before and after the relative radiometric correction of the night-light image in the embodiment of the present invention, where (a) and (c) are before the radiometric correction, and (b) and (d) are after the radiometric correction. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] The technical solutions of the present invention will be described in detail below with reference to the drawings and embodiments.

[0055] After the satellite is launched, due to the changes in the space environment it is subjected to, such as temperature changes, physical vibrations, etc., and the attenuation of the stability of the detector elements of the satellite sensor itself, the response model of the satellite sensor detector elements changes, resulting in the unusability of the calibration results before satellite launch, and on-orbit relative radiometric calibration must be carried out. For the night-light remote sensing satellite to capture the weak light on the earth's surface and the weak light signal scattered by the moon, the existing conventional on-orbit radiometric calibration methods cannot be directly applied to the night-light remote sensing satellite. Based on the imaging characteristics of the night-light remote sensing satellite, that is, there is a certain model relationship between the radiation responses of its different imaging parameters, by establishing a radiation response model between different imaging parameters, the association between the daytime radiation reference and the night-time imaging images can be achieved; a radiation reference transfer model from daytime to night-time for the night-light remote sensing satellite is constructed, and the calibration of the night-time image data of the night-light remote sensing satellite is achieved by using the daytime calibration imaging data of the night-light remote sensing satellite, and finally the on-orbit relative radiometric calibration of the night-light remote sensing satellite is completed.

[0056] A method for on-orbit relative radiometric calibration of a night-light remote sensing satellite based on radiation reference transfer provided by an embodiment of the present invention specifically includes the following steps:

[0057] Step 1, extraction of the daytime radiation reference. Based on the imaging data of the night-light remote sensing satellite for large-area uniform ground objects with different reflectivities during the day, a radiometric calibration model is established to extract the daytime radiation reference, and this reference is manifested as the relative radiation coefficients of each detector element of the night-light remote sensing satellite. Specifically:

[0058] Step 1.1, Daytime calibration imaging of the night light remote sensing satellite. According to the actual resolution of the night light remote sensing satellite, large-area uniform ground objects with different reflectivities that meet the following requirements are selected globally to obtain daytime calibration imaging data.

[0059] Condition (1): The brightness of the large-area uniform ground objects with different reflectivities can cover at least 3 different brightness levels in the dynamic range of the night light sensor. When the on-orbit radiation calibration model of the night light remote sensing satellite is in the linear mode, if the surface uniform field data with different reflectivities only shows 1 brightness level in the imaging dynamic range of the night light remote sensing satellite, then b in formula (1) is 0, and only the coefficient a is solved.

[0060] Condition (2): The large-area uniform ground objects need to cover all the detector elements of the night light remote sensing satellite;

[0061] Condition (3): The non-uniformity of the uniform scene data needs to be higher than the calibration accuracy, otherwise the requirement of the calibration accuracy cannot be achieved.

[0062] Brightness level: It refers to the minimum quantization unit of the remote sensing satellite. For example, if 8-bit quantization is adopted, there are 256 brightness levels;

[0063] Non-uniformity: It refers to the non-uniform degree of the uniform ground object itself, which is measured by the ratio of the standard deviation to the mean value of the uniform area; for example, if the calibration accuracy requirement is better than 3%, then the non-uniformity of the uniform ground object used for calibration should be better than 0.3%. Otherwise, it is difficult to achieve the calibration accuracy requirement of 3% using the data of this uniform ground object.

[0064] Step 1.2, Construct the on-orbit radiation calibration model of the night light remote sensing satellite as follows:

[0065]

[0066] In the formula, j is the number of large-area uniform fields with different reflectivities, and DN i,j is the radiation quantization value (also known as the image gray value) of the i-th detector element of the night light remote sensing satellite in the j-th uniform field, is the gray mean value of all imaging detector elements of the night light remote sensing satellite in the j-th uniform field, a i and b i are the relative radiation coefficients of the i-th detector element of the night light remote sensing satellite. Note: This model can be rewritten as a non-linear model such as a second-order or third-order polynomial model, with the same principle.

[0067] Step 1.3, For the i-th imaging detector element of the night light remote sensing satellite, based on the daytime calibration image data obtained in Step 1.1, use the least squares method to solve formula (1) to obtain the relative radiation coefficients a i and b i .

[0068] Step 1.4, repeat Step 1.3 to obtain the relative radiation coefficients a i and b i (i ∈ [0, M], where M is the number of detector elements of the night light remote sensing satellite), that is, the daytime radiation benchmark is extracted.

[0069] Step 2, construct a radiation transfer model between different imaging parameters of the night light remote sensing satellite, specifically as follows.

[0070] Step 2.1, according to Step 1, establish a radiation calibration model for the night light remote sensing satellite under different imaging parameters. Different imaging parameters include low gain, high gain, low level, and high level. Here, the different gain parameters are taken as an example for illustration. According to Equation (I), when the night light remote sensing satellite images under high gain parameter (HighGain, hereinafter referred to as H) and low gain parameter (LowGain, hereinafter referred to as L), its calibration models are respectively:

[0071]

[0072] In the formula, a H , b H , a L , b L correspond to the radiation coefficients under high and low gains respectively, DN H , DN L are the gray values of the high and low gain images of the night light remote sensing satellite, is the average gray value of the high and low gain images;

[0073] Step 2.2, define the radiation response model relationship between different imaging parameters of the night light remote sensing satellite. Here, the different gain parameters are taken as an example for illustration. According to the imaging characteristics of the night light remote sensing satellite, there is a certain model relationship between the radiation responses of its different imaging parameters. For example, there is an unknown order polynomial model relationship between the gray value of the high gain image and the gray value of the low gain image,

[0074] DN h = B0 + B1 × DN L + B2 × DN L 2 +…+ B n × DN L n (IV)

[0075] Furthermore, the average gray values of different imaging images also have,

[0076]

[0077] In the formula, n is the order of the polynomial, n is greater than or equal to 1; B0, B1, B2…B nis the model relationship coefficient;

[0078] In specific implementation, the value of n varies with different satellite designs and can be solved based on the calibration data before the satellite launch or the in-orbit data.

[0079] For example, the order n can be determined by any of the following two methods:

[0080] 1. Based on satellite design values;

[0081] 2. Optimally fit the satellite imaging data and select the polynomial order that minimizes the sum of squared fitting residuals.

[0082] Step 2.3, combine equations (II) to (V) to establish the grayscale value DN of the image with different imaging parameters of the night light remote sensing satellite H DN L The model relationship between them can realize the radiation transfer between images with arbitrary imaging parameters of night light remote sensing satellites.

[0083] The specific implementation is to take high and low gain images as an example to illustrate. This step converts a into H 、b H Eliminate and establish DN H and The model relationship, through a in this model L 、b L That is, the radiation transfer from low-gain image to high-gain image is realized. For example, when n = 2, the conversion model of low-gain image to high-gain image of night light remote sensing satellite is as follows:

[0084]

[0085] When n>2, DN H and Just solve the relational combination formulas (2)-(5).

[0086] In step 3, based on the radiation transfer model in 2.3, the extracted daytime radiation benchmark of the night light remote sensing satellite can be transferred to the nighttime to obtain the on-orbit relative radiation coefficient of the night light remote sensing satellite.

[0087] The specific implementation is to take high and low gain images as an example to illustrate. In step 1, the relative radiation coefficient a of each detector of the night light remote sensing satellite under the daytime imaging parameters (such as low gain parameters) has been calculated by using the daytime radiation calibration data. i and b i , that is, the daytime radiation benchmark. Based on the radiation transfer model between different imaging parameters of night light remote sensing satellites established in step 2, the daytime radiation benchmark can be transferred to the nighttime. Let a L =a i 、b L =bi , substitute into formula (VI), that is, transfer the daytime radiation reference a i and b i to night. For example, when n = 2, the relative radiation coefficients a L and b L of each detector element of the night light remote sensing satellite to night are as follows.

[0088]

[0089] When n>2, solve by combining formulas (II)-(V).

[0090] Step 4, extract image samples of different imaging parameters from the laboratory radiometric calibration image data or on-orbit imaging data before the launch of the night light remote sensing satellite (the laboratory radiometric calibration image data is not less than 50 scenes, and the on-orbit imaging data is not less than 100 scenes), such as (DN H , DN L ), and solve the model coefficients B0, B1, B2... B n .

[0091] Step 5, substitute the daytime radiation coefficients a i and b i of the night light remote sensing satellite obtained in Step 1 and the model coefficients of formula (IV) obtained in Step 4 into Step 3 to realize the relative radiometric correction of the night light remote sensing satellite's night image data and complete the on-orbit relative radiometric calibration of the night light remote sensing satellite.

[0092] In specific implementation, the above process can be automatically run using computer software technology. The hardware for running and implementing this method should also be within the protection scope of the present invention.

[0093] On the other hand, the embodiment of the present invention also provides an on-orbit relative radiometric calibration system for a night light remote sensing satellite based on radiation reference transfer, including:

[0094] A processor and a memory, the memory is used to store program instructions, and the processor is used to call the stored instructions in the memory to execute the on-orbit relative radiometric calibration method for a night light remote sensing satellite based on radiation reference transfer as described in the above technical solution.

[0095] In a specific embodiment, five uniform field data with different reflectivity magnitudes are used to construct a radiometric calibration model and a fitting residual level of the night light remote sensing satellite, such as Figure 1 taking the linear model as an example. Using the on-orbit data of the night light remote sensing satellite, a radiation transfer model between different imaging parameters of the night light remote sensing satellite is established, such as Figure 2Taking a second-order polynomial as an example. The radiation reference extracted from the daytime radiation calibration model is passed through the radiation transfer model to achieve relative radiation correction of the night-time image data of the night-light remote sensing satellite. The fringe coefficient is used as an objective evaluation index to evaluate the night-light image after relative radiation correction. The objective evaluation index is shown in Table 1, and the visualization correction effect comparison is as Figure 3 shown. It can be seen from this that the method of the present invention improves the quality of the night-light satellite image.

[0096] Table 1 Statistical table of relative correction accuracy of night-light images

[0097]

[0098] The specific embodiments described in this article are only illustrative of the spirit of the present invention. Those skilled in the art of the present invention can make various modifications or supplements to the described specific embodiments or use similar methods to replace them, but will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. A method for on-orbit relative radiometric calibration of a night light remote sensing satellite based on radiation reference transfer, characterized in that, Including the following steps: Step 1: Obtain daytime calibration imaging data and perform daytime radiation benchmark extraction. The daytime radiation benchmark is manifested as the relative radiation coefficients of each detector element of the night light remote sensing satellite. Step 2: Construct a radiation transfer model between different imaging parameters of the night light remote sensing satellite to achieve radiation transfer between images with any imaging parameters of the night light remote sensing satellite. Step 3: Based on the radiation transfer model in Step 2, transfer the extracted daytime radiation benchmark of the night light remote sensing satellite to nighttime to obtain the on-orbit relative radiation coefficients of the night light remote sensing satellite. Step 4: Extract image samples with different imaging parameters from the laboratory radiation calibration image data before the launch of the night light remote sensing satellite or on-orbit imaging data, and solve the relationship coefficients of the radiation transfer model. Step 5: Substitute the daytime radiation coefficients of the night light remote sensing satellite obtained in Step 1 and the relationship coefficients of the radiation transfer model obtained in Step 4 into the radiation transfer model in Step 3 to achieve relative radiation correction of the night light remote sensing satellite's nighttime image data, and complete the on-orbit relative radiation calibration of the night light remote sensing satellite.

2. The method for on-orbit relative radiometric calibration of a nocturnal remote sensing satellite based on radiation reference transfer according to claim 1, wherein: In Step 1, according to the actual resolution of the night light remote sensing satellite, select large-area uniform ground objects with different reflectivities that meet the following requirements globally to obtain daytime calibration imaging data. The large-area uniform ground objects should meet the following conditions: Condition (1): The brightness of the large-area uniform ground objects with different reflectivities can cover at least 3 different brightness levels in the dynamic range of the night light sensor. Condition (2): The large-area uniform ground objects cover all imaging detector elements of the night light remote sensing satellite. Condition (3): The non-uniformity of the uniform field data used for daytime calibration itself should be higher than the calibration accuracy, otherwise the requirements of the calibration accuracy cannot be met.

3. The method for on-orbit relative radiometric calibration of a nocturnal remote sensing satellite based on radiation reference transfer according to claim 1, wherein: The specific implementation of daytime radiation benchmark extraction includes: First, construct an on-orbit radiation calibration model for the night light remote sensing satellite. The on-orbit radiation calibration model for the night light remote sensing satellite is a first-order linear model or a multi-order non-linear model. Then, for the i-th imaging detector element of the night light remote sensing satellite, based on the obtained daytime calibration image data, use the least squares method to solve the on-orbit radiation calibration model of the night light remote sensing satellite to obtain the relative radiation coefficient of the i-th imaging detector element, where i ∈ [0, M], and M is the number of detector elements of the night light remote sensing satellite. Finally, obtain the relative radiation coefficients of all detector elements of the night light remote sensing satellite, that is, extract the daytime radiation benchmark.

4. The method for on-orbit relative radiometric calibration of a night-light remote sensing satellite based on radiation reference transfer according to claim 3, wherein: When the on-orbit radiation calibration model of the night light remote sensing satellite is a first-order linear model, its calculation formula is as follows: In the formula, j is the number of large - area uniform fields with different reflectivities, and DN i,j is the radiation quantization value of the i - th detector element of the j - th uniform - field night - light remote - sensing satellite, is the gray - scale mean value of all imaging detector elements of the j - th uniform - field night - light remote - sensing satellite, a i and b i are the relative radiation coefficients of the i - th detector element of the night - light remote - sensing satellite.

5. The method for on-orbit relative radiometric calibration of a night-light remote sensing satellite based on radiation reference transfer according to claim 1, characterized in that: The specific implementation method of Step 2 is as follows; Step 2.1: According to Step 1, establish radiation calibration models under different imaging parameters of the night light remote sensing satellite. When the night light remote sensing satellite images under different imaging parameters, its calibration models are respectively: where a H , b H , a L , b L correspond to the radiation coefficients under different imaging parameters, DN H , DN L are the image gray values of the night light remote sensing satellite under different imaging parameters, is the average image gray value under different imaging parameters, and the subscripts H and L are different imaging parameters, where different imaging parameters include low gain, high gain, low level, and high level; Step 2.2: Define the radiation response model relationship between different imaging parameters of the night light remote sensing satellite. According to the imaging characteristics of the night light remote sensing satellite, there is a certain model relationship between the radiation responses of its different imaging parameters, and there is an unknown-order polynomial model relationship between the image gray values. DN H = B0 + B1 × DN L + B2 × DN L 2 + … + B n × DN L n (4) Furthermore, the gray-scale means of different imaging images also have: where n is the polynomial order and n is greater than or equal to 1; B0, B1, B2…B n are the model relationship coefficients; Step 2.3, combining equations (2)-(5), the relationship between the gray values DN H and DN L of different imaging parameter images of the night light remote sensing satellite can be established, and the radiation transfer between any imaging parameter images of the night light remote sensing satellite can be realized.

6. The method for on-orbit relative radiometric calibration of a night light remote sensing satellite based on radiation reference transfer according to claim 5, wherein: The specific implementation method of Step 2.3 is as follows: Eliminate a H and b H through a polynomial relationship, and establish the model relationship between DN H and . The radiative transfer of images with different imaging parameters is realized through a and b L and b L in the model; when n = 2, the conversion model of images of the night light remote sensing satellite under different imaging parameters at night is as follows: When n > 2, DN H and can be solved by combining formulas (2) - (5).

7. The method for on-orbit relative radiometric calibration of a nocturnal remote sensing satellite based on radiation reference transfer according to claim 5, wherein: Determine the order n by any one of the following two methods: (1) Based on the satellite design value; (2) Based on the optimal fitting of satellite imaging data, take the polynomial order when the sum of the squares of the fitting residuals is the smallest.

8. The method for on-orbit relative radiometric calibration of a night light remote sensing satellite based on radiation reference transfer according to claim 6, wherein: The specific implementation method of Step 3 is as follows: In Step 1, the relative radiation coefficients a i and b i of each detector element of the night light remote sensing satellite under the daytime imaging parameters have been calculated through the daytime radiometric calibration data, that is, the daytime radiation benchmark, where i ∈ [0, M] and M is the number of detector elements of the night light remote sensing satellite; Based on the radiative transfer model between different imaging parameters of the night-light remote sensing satellite established in Step 2, the daytime radiation benchmark can be transferred to the night; let a L = a i , b L = b i , substitute into formula (VI), that is, transfer the daytime radiation benchmarks a i , b i to the night; when n = 2, the transfer models of the relative radiation coefficients a L , b L of each detector element of the night-light remote sensing satellite to the night are as follows: When n > 2, solve it by combining formulas (2)-(5).

9. The method for on-orbit relative radiometric calibration of a night light remote sensing satellite based on radiation reference transfer according to claim 1, characterized in that: In Step 4, the laboratory radiometric calibration image data is not less than 50 scenes, and the on-orbit imaging data is not less than 100 scenes.

10. A relative radiometric calibration system for on-orbit night remote sensing satellites based on radiation reference transfer, characterized in that, Including: A processor and a memory. The memory is used to store program instructions, and the processor is used to call the stored instructions in the memory to execute the on-orbit relative radiometric calibration method of the night light remote sensing satellite based on radiation reference transfer as described in any one of claims 1-9.

Images