A method and device for removing stray light from a twilight orbit satellite's micro-light channel

Abstract

The embodiment of the present application discloses a method and device for removing stray light from the low-light channel of a twilight orbit satellite. The method includes: obtaining the original image of MERSI / LLB; processing it with a stray light removal algorithm. The algorithm includes: performing sliding fitting on the original amplitude brightness value R1 scan line by scan line to obtain the fitted radiance F1; extracting continuous light pixels according to the difference between R1 and F1; performing linear fitting on the radiance of the continuous light pixels, and replacing R1 with the fitting result to obtain the radiance value R2; performing sliding fitting on R2 to obtain the fitted radiance F2; and obtaining the result after removing stray light according to the values ​​of R1, F2 and background radiance. The advantages of the present application are as follows: (1) It is simple and easy to implement, and does not require the advance calculation of a series of parameters for estimating stray light; (2) It is not affected by seasonal changes; (3) It can remove stray light that changes between frames, which cannot be achieved by the original solution.

Description

Technical Field

[0001] The present invention relates to the technical field of satellite meteorology, and in particular to a method and a device for removing stray light from a micro-light channel of a dawn-dusk orbit satellite. Background Art

[0002] Low-light remote sensing is of great significance for observing and understanding the climate system. It can be used to analyze weather cloud maps, detect the extent of sea ice, the turbidity of seawater, detect fire points, volcanic ash plumes and pyroclastic flows, etc. Compared with other satellite data, low-light data is unique in that it can directly reflect human activities, such as identifying ships to combat illegal fishing, identifying natural gas flaring to assess greenhouse gas emissions, and studying city-related issues through night lights. Therefore, low-light remote sensing has gradually become a research hotspot in the field of remote sensing.

[0003] At present, the main low-light detection data in the world comes from the Day / Night Band (DNB) channel of the Visible Infrared Imaging Radiometer Suite (VIIRS) carried by the US SNPP and NOAA-20 satellites. VIIRS / DNB has 16 detectors, a spatial resolution of 750m, a width of about 3000km, and three gain stages of low, medium and high. It can image under daytime sunlight conditions, dawn and dusk conditions, and low illumination conditions at night, thereby achieving a large dynamic range of radiance observations of 7 orders of magnitude. The low-gain stage can be calibrated on the satellite, and the calibration of the medium and high gain stages can be achieved through calibration transfer. SNPP and NOAA-20 are both afternoon satellites. They pass through the local time at about 1:30 a.m. when descending, crossing the terminator at the North and South Poles. VIIRS / DNB is seriously polluted by stray light near the terminator, which greatly reduces the image quality. The so-called stray light refers to the radiation emitted or reflected by non-target objects received by the satellite. There are several possible reasons for its generation:

[0004] 1. Optical mirror scattering due to rough or unclean mirror surface;

[0005] 2. The light scattered by the shading plate enters the optical path of the detector;

[0006] 3. Light leakage caused by holes or gaps in the internal structure of the instrument.

[0007] Take the data of VIIRS / DNB on June 10, 2012 as an example (e.g. Figure 1aAs shown in the figure, stray light creates a gray mask on the upper part of the image. The VIIRS image product development team found that DNB stray light is related to the following factors: detector number, scanning angle (i.e. pixel number), the angle of sunlight relative to the subsatellite normal vector (referred to as the solar incidence angle), the azimuth of the sun relative to the subsatellite normal vector (referred to as the solar azimuth angle), and the different faces of the half-angle mirror (the half-angle mirror has two faces, A and B). Based on this, their stray light removal solution is as follows:

[0008] (1) Filter out dark pixels: remove pixels with solar zenith angle or lunar zenith angle less than 105°. In addition, use the 2003 World Night Lights Dataset to remove pixels with lights.

[0009] (2) Grouping dark pixels: The data of each half-angle mirror of each detector are grouped according to the scanning angle. The 4064 pixels on a scan line are divided into 127 groups, with 32 pixels in each group. In order to reduce noise, the median of the radiance of these 32 pixels is taken, and the average angle of incidence of sunlight is recorded at the same time.

[0010] (3) Fitting the relationship between the incident angle of sunlight and the radiant brightness: Taking the 108th group of pixels (at the right end of the scanning line) on the B surface of the half-angle mirror at 11:00 on June 18, 2021 as an example, Figure 1b The relationship between the radiance received by 16 detectors and the cosine of the solar incidence angle (cosθ) is shown. cosθ=0.27 corresponds to the dawn-dusk zone. As cosθ increases, sunlight directly enters the detector, resulting in enhanced radiance. When cosθ=0.45, sunlight enters the earth's penumbra, and the radiance rapidly decreases to a basically stable background value. The background value of radiance is caused by airglow. Therefore, the intensity of stray light is obtained by subtracting the background value from the satellite received radiance. The stray light is fitted in sections. A quadratic polynomial is used to fit the relationship between radiance and cosθ in the radiance rising interval, and a quartic polynomial is used to fit the earth's penumbra. The radiance weakening interval at the far left needs to be obtained by extrapolation.

[0011] (4) The above three steps obtain the relationship between stray light and the incident angle of sunlight for each detector under different half-angle mirror states and different scanning angles. Therefore, the stray light size of each pixel can be estimated by the detector number, half-angle mirror state, pixel number and incident angle of sunlight, and then the stray light is removed by subtracting the original radiance. Repeat steps (1) to (3) using the data of each month to remove the stray light of that month, thereby solving the impact caused by the change of the solar azimuth angle.

[0012] On July 5, 2021, my country's second-generation polar-orbiting meteorological satellite Fengyun-3E (FY-3E) was successfully launched. It is the first dawn-dusk orbit satellite in the Fengyun satellite family and the world's first dawn-dusk orbit operational meteorological satellite. Its descending transit time is about 5:40 local time and the ascending orbit time is about 17:40. The networking with the morning and afternoon satellites has achieved full coverage of global data every 6 hours, meeting the requirements of the 6-hour assimilation time window of numerical weather forecasting. The payload medium-resolution imager (MERSI) of FY-3E is equipped with a low-light band (LLB) with a subsatellite resolution of 1km and a width of 2500km. LLB has three gain stages of low, medium and high, and can observe radiances of 3.5×10-5 to 50W / m2 / sr, making the Fengyun series of satellites have the ability to detect low light at night for the first time. FY-3E is always near the terminator. MERSI was designed with the particularity of the terminator orbit in mind. To avoid the influence of stray light, the area of ​​the sunshade was increased and the swath of the scanning mirror on the day side was reduced. However, unexpectedly, the data of LLB on the night side was seriously polluted by stray light. Preliminary analysis found that the stray light characteristics of MERSI / LLB are very different from those of VIIRS / DNB. Specifically, the stray light on the same scan line varies greatly with the scan angle; the stray light varies complexly with the solar incident angle; and the intensity of the stray light varies periodically between frames. Applying the stray light removal scheme of VIIRS / DNB to MERSI / LLB requires reducing the number of pixels in each group in step (2) and making more segments in step (3) to fit the relationship between radiance and solar incident angle. This greatly increases the amount of calculation, and most importantly, the stray light that varies between MERSI / LLB frames cannot be removed. Summary of the invention

[0013] Since the stray light of MERSI / LLB varies with the scanning angle and the solar incident angle in a complex manner and has periodic inter-frame variations, the stray light removal scheme of VIIRS / DNB is not effective for MERSI / LLB. The purpose of the embodiment of the present application is to provide a stray light removal method based on the characteristics of MERSI / LLB stray light, which can better remove the stray light of MERSI / LLB.

[0014] To achieve the above objectives, in a first aspect, an embodiment of the present invention provides a method for removing stray light from a twilight orbit satellite micro-light channel, comprising:

[0015] Acquire MERSI / LLB original images; the original images only include pure nighttime data;

[0016] Using a stray light removal algorithm to process the image to be processed to obtain a target image;

[0017] The stray light removal algorithm comprises the following steps:

[0018] Perform sliding fitting on the original image scan line by scan line, assuming that the amplitude brightness value of a certain scan line is R1, and the fitting radiance value is F1;

[0019] Extract the pixel number and radiance of the first and last pixels of the continuous light pixel according to the difference between the amplitude brightness value R1 and the fitting amplitude brightness F1;

[0020] Perform linear fitting on the radiance of continuous light pixels to obtain fitting results, and use the fitting results to replace the amplitude brightness value R1 of the corresponding pixel to obtain a new amplitude brightness value R2;

[0021] Perform sliding fitting on the new amplitude brightness value R2 to obtain fitting radiance F2;

[0022] According to the amplitude brightness value R1, the fitting radiance F2 and the preset background radiance value, the target image after removing the stray light is obtained.

[0023] In certain preferred embodiments of the present application, before acquiring the MERSI / LLB original image, the method further includes:

[0024] The solar zenith angle data of MERSI / LLB is read, and pure night data is selected therefrom; the pure night data is data with a solar zenith angle greater than 101°.

[0025] As a specific implementation of the present application, the amplitude brightness value R1 is fitted to obtain the fitted radiance F1, specifically:

[0026] Use Savitzky-Golay filter, take 61 pixels as window, and use cubic polynomial to make sliding fit to the amplitude brightness value R1 to get the fitting radiance F1; or

[0027] The amplitude brightness value R1 is subjected to sliding fitting by sliding average, Fourier filtering or empirical mode decomposition to obtain the fitted radiance F1.

[0028] As a specific implementation of the present application, the pixel number and radiance of the first and last pixels of the continuous light pixel are extracted according to the amplitude brightness value R1 and the fitting amplitude brightness F1, specifically:

[0029] Calculate the difference between the fitting amplitude brightness F1 and the amplitude brightness value R1, take the absolute value of the difference and perform 11-point sliding average to obtain a sliding average result;

[0030] The sliding average result is judged by using a set threshold, pixels that are continuously greater than the threshold are judged as continuous light pixels, and the pixel numbers and radiances of the first and last pixels are extracted.

[0031] As a specific implementation of the present application, a linear fitting is performed on the radiance of the continuous light pixel to obtain a fitting result, which is specifically:

[0032] The pixel numbers of the extracted first and last pixels are taken as the independent variables, and their radiance is taken as the dependent variable to perform linear fitting, and the radiance of the pixels between the first and last pixels is fitted to obtain the fitting result.

[0033] As a specific implementation of the present application, the new amplitude brightness value R2 is fitted to obtain the fitted radiance F2, specifically:

[0034] Using a Savitzky-Golay filter to fit the new amplitude brightness value R2 to obtain a fitting radiance F2; or

[0035] The new amplitude brightness value R2 is subjected to sliding fitting by adopting sliding average, Fourier filtering or empirical mode decomposition to obtain the fitting radiance F2.

[0036] As a specific implementation of the present application, according to the amplitude brightness value R1, the fitting radiance F2 and the preset background radiance value, the target image after removing the stray light is obtained, specifically:

[0037] Calculate the difference between the amplitude brightness value R1 and the fitted radiance F2, and add the background radiance value caused by airglow to get the result of removing stray light;

[0038] The background radiance value caused by airglow is determined by MERSI / LLB observations in the Antarctic region during June and July without any spurious nights.

[0039] In the second aspect, an embodiment of the present application also provides a device for removing stray light from the micro-light channel of a dawn-dusk orbit satellite, comprising a processor, an input device, an output device and a memory, wherein the processor, input device, output device and memory are interconnected, wherein the memory is used to store a computer program, and the computer program includes program instructions, and the processor is configured to call the program instructions to execute the method of the above-mentioned first aspect.

[0040] In a third aspect, the embodiment of the present application further provides another device for removing stray light from a twilight orbit satellite micro-light channel, comprising:

[0041] An acquisition unit, used for acquiring a MERSI / LLB original image; the original image only includes pure nighttime data;

[0042] A processing unit, configured to process the original image using a stray light removal algorithm to obtain a target image;

[0043] The stray light removal algorithm comprises the following steps:

[0044] Perform sliding fitting on the original image scan line by scan line, assuming that the amplitude brightness value of a certain scan line is R1, and the fitting radiance is F1;

[0045] Extract the pixel number and radiance of the first and last pixels of the continuous light pixel according to the difference between the amplitude brightness value R1 and the fitting amplitude brightness F1;

[0046] Perform linear fitting on the radiance of continuous light pixels to obtain fitting results, and use the fitting results to replace the amplitude brightness value R1 of the corresponding pixel to obtain a new amplitude brightness value R2;

[0047] Fitting the new amplitude brightness value R2 to obtain a fitting radiance F2;

[0048] According to the amplitude brightness value R1, the fitting radiance F2 and the preset background radiance value, the target image after removing the stray light is obtained.

[0049] The implementation of the present application has the following advantages:

[0050] 1. Compared with the existing stray light removal solution, this application is simpler and easier to implement, and does not require the advance calculation of a series of parameters for estimating stray light.

[0051] 2. The original stray light removal scheme updates the parameters for estimating stray light once a month. Since the solar azimuth changes continuously within a month, the accuracy of the estimated stray light will change. However, the scheme of the present application is not affected by seasonal changes, and theoretically the accuracy of removing stray light remains unchanged.

[0052] 3. The solution of the present application can remove stray light that varies between frames, while the original solution cannot achieve this. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the specific implementation of the present invention or the technical solution in the prior art, the drawings required for use in the specific implementation or the description of the prior art are briefly introduced below.

[0054] Figure 1a This is a VIIRS / DNB data image from June 10, 2012;

[0055] Figure 1b This is the relationship between the radiation intensity of different detectors (indicated by thin lines of different colors) of the 108th group of pixels on the B surface of the half-angle mirror and the cosine of the solar incident angle at 11:00 on June 18, 2021;

[0056] Figure 2aThis is a schematic diagram of the radiance (scatter points) of a scan line passing through Beijing at 22:15 UTC on January 2, 2022, and the sliding fitting value of the radiance using the Savitzky-Golay filter (red line);

[0057] Figure 2b yes Figure 2a Schematic diagram of the absolute value of the difference between the medium radiance and the fitted value (scatter points) and its eleven-point sliding average (red line);

[0058] Figure 2c It is a schematic diagram of the radiance (scatter points) after removing the light pixels and its sliding fitting value (red line);

[0059] Figure 2d is a schematic diagram of the radiance after removing the basic radiance;

[0060] Figure 3 This is a flow chart of a method for removing stray light from a twilight orbit satellite micro-light channel provided in an embodiment of the present application;

[0061] Figure 4a This is the original image of MERSI / LLB at 22:15UTC on January 2, 2022. The contour lines represent the solar altitude angle;

[0062] Figure 4b yes Figure 4a After removing stray light, the image only shows the pure night area.

[0063] Figure 5 It is a schematic diagram of the structure of a stray light removal device for a dawn-dusk orbit satellite micro-light channel provided in an embodiment of the present application. DETAILED DESCRIPTION

[0064] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0065] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0066] To better understand the embodiments of the present invention, the following takes the pure night data of a scan line passing through Beijing at 22:15 UTC on January 2, 2022 as an example to explain the principle of our method:

[0067] Figure 2a The horizontal axis in is the serial number of the pixel on the scan line. The solar zenith angle decreases as the serial number increases. The blue scattered points are the radiance on the scan line, in units of W / m2 / sr. It can be seen that the radiance has a trend of increasing with the increase of the solar zenith angle. This growth trend is mainly affected by stray light, which is the object to be removed in the method of this embodiment.

[0068] in addition, Figure 2a The 200-270, 630-700, and 770-840 pixels are affected by city lights, and their radiance values ​​are significantly higher than those of the surrounding pixels. Based on this, the radiance can be divided into two parts: the basic radiance and the radiance of city lights. The basic radiance can also be divided into two parts: one is the stray light with a trend change, and the other is the constant background value caused by airglow.

[0069] In order to obtain the changing trend of the basic radiance, we use the Savitzky-Golay filter to do sliding fitting ( Figure 2a However, the fitting result at the light pixel will obviously deviate from the basic radiation value, so it is necessary to remove the light and do the fitting again. The difference between the radiance of the light pixel and the fitting value is large ( Figure 2b In order to eliminate the influence of individual dark pixels in the city, we make a sliding average of the absolute value of the difference between the two ( Figure 2b The middle curve), the pixels larger than the threshold are judged as light pixels.

[0070] After selecting the light pixels, linear fitting is performed on the starting and ending pixels with the pixel number as the independent variable and the basic radiance value as the dependent variable, and the original radiance value is replaced by the fitted radiance value ( Figure 2c scatter points in the middle) to remove the radiance of the light.

[0071] Then use the Savitzky-Golay filter again to do a sliding fit on the radiance of the removed light ( Figure 2c The smoothed basic radiance is obtained, and then the original radiance value is subtracted from the basic radiance ( Figure 2d ), thus removing stray light and background radiance.

[0072] Finally, a constant value representing the airglow radiance is added to obtain the radiance without stray light. The airglow radiance value is determined by MERSI / LLB observations in the Antarctic region during June and July without stray light.

[0073] Based on the above algorithm principle, this application provides a method for removing stray light from the dusk-dawn orbit satellite low-light channel, which is mainly aimed at dealing with the problem that the data on the night side of the LLB is seriously polluted by stray light. Figure 3 As shown, the stray light removal method mainly includes the following steps:

[0074] Acquire MERSI / LLB original images; the original images only include pure nighttime data;

[0075] Processing the original image using a stray light removal algorithm to obtain a target image;

[0076] The stray light removal algorithm comprises the following steps:

[0077] Sliding fitting is performed on each scan line of the original image, assuming that the amplitude brightness value of a certain scan line is R1, and the fitting radiance is F1; that is, sliding fitting is performed on each scan line;

[0078] Extract the pixel number and radiance of the first and last pixels of the continuous light pixel according to the difference between the amplitude brightness value R1 and the fitting amplitude brightness F1;

[0079] Perform linear fitting on the radiance of continuous light pixels to obtain fitting results, and use the fitting results to replace the amplitude brightness value R1 of the corresponding pixel to obtain a new amplitude brightness value R2;

[0080] Fitting the new amplitude brightness value R2 to obtain a fitting radiance F2;

[0081] According to the amplitude brightness value R1, the fitting radiance F2 and the preset background radiance value, the target image after removing the stray light is obtained.

[0082] As a specific implementation, the specific steps of the stray light removal algorithm are as follows:

[0083] (1) Read the MERSI / LLB radiance data and solar zenith angle, select pure night data (solar zenith angle>101°), set the original radiance value of one of the scan lines as R1, use the Savitzky-Golay filter, take 61 pixels as the window, and perform sliding fitting of R1 with a cubic polynomial to obtain the fitted radiance F1.

[0084] (2) Subtract F1 from R1 and take the absolute value to make an 11-point sliding average. Set a threshold for the sliding average result and extract the first and last pixel numbers for pixels that are continuously greater than the threshold.

[0085] (3) A linear fit is performed with the pixel numbers of the first and last pixels extracted as the independent variable and their radiant brightness as the dependent variable. The radiant brightness of the pixels between the first and last pixels is fitted, and the fitting result is used to replace the R1 value of the corresponding pixel to obtain R2.

[0086] (4) Use the Savitzky-Golay filter to fit R2 and obtain the fitted radiance F2.

[0087] (5) Subtract F2 from R1, and add the background radiance value caused by airglow to obtain the result of removing stray light.

[0088] It should be noted that since there are many ways to filter, the methods for fitting radiance in step 1 and step 4 are not unique, such as using sliding average, Fourier filtering, empirical mode decomposition, etc., but their essence is to extract the low-frequency change trend of radiance. There are also many ways to extract light pixels in step 2, such as judging by the severity of radiance changes. The method for obtaining the basic radiance of light pixels in step 3 is also not unique. You can use polynomial fitting or directly assign the radiance values ​​of the first and last pixels.

[0089] Please refer to Figure 4a and Figure 4b , is a graph showing the effect of the algorithm in this application. Take a 5-minute segment of MERSI / LLB data when UTCFY-3E was over my country at 22:15 on January 2, 2022 as an example to verify the stray light removal algorithm ( Figure 4a ), from the solar altitude angle, we can see that the right side of the image is the morning and evening area, and the left side is the night. The bright white area on the left side is affected by stray light. Figure 4b As shown (only pure night data is shown), it can be seen that the bright white area on the left has been removed, and some light pixels have been displayed, and the effect is basically satisfactory.

[0090] From the above description, it can be seen that the method for removing stray light from the twilight channel of a dawn-dusk orbit satellite provided in the embodiment of the present application has the following advantages:

[0091] 1. Compared with the existing stray light removal solution, this application is simpler and easier to implement, and does not require the calculation and estimation of a series of stray light parameters in advance.

[0092] 2. The original stray light removal scheme updates the parameters for estimating stray light once a month. Since the solar azimuth changes continuously within a month, the accuracy of the estimated stray light will change. However, the scheme of the present application is not affected by seasonal changes, and theoretically the accuracy of removing stray light remains unchanged.

[0093] 3. The solution of the present application can remove stray light that varies between frames, while the original solution cannot achieve this.

[0094] Based on the same inventive concept, the embodiment of the present application provides a device for removing stray light from a twilight orbit satellite micro-light channel, comprising:

[0095] An acquisition unit, used for acquiring a MERSI / LLB original image; the original image only includes pure nighttime data;

[0096] A processing unit, configured to process the original image using a stray light removal algorithm to obtain a target image;

[0097] The stray light removal algorithm comprises the following steps:

[0098] (1) performing sliding fitting on the original image scan line by scan line, assuming that the amplitude brightness value of a certain scan line is R1 and the fitting radiance is F1;

[0099] (2) extracting the pixel number and radiance of the first and last pixels of the continuous light pixel according to the original amplitude brightness value R1 and the fitted amplitude brightness F1;

[0100] (3) performing linear fitting on the radiance of the continuous light pixel to obtain a fitting result, and using the fitting result to replace the original amplitude brightness value R1 of the corresponding pixel to obtain a new amplitude brightness value R2;

[0101] (4) fitting the new amplitude brightness value R2 to obtain the fitting radiance F2;

[0102] (5) According to the original amplitude brightness value R1, the fitted radiance F2 and the preset background radiance value, a target image after removing stray light is obtained.

[0103] Wherein, step (1) is specifically as follows:

[0104] Use Savitzky-Golay filter, take 61 pixels as window, and do sliding fitting of the original amplitude brightness value R1 with cubic polynomial to get the fitting radiance F1; or

[0105] The sliding fitting is performed on the original amplitude brightness value R1 by using sliding average, Fourier filtering or empirical mode decomposition to obtain the fitting radiance F1.

[0106] Wherein, step (2) is specifically as follows:

[0107] Calculate the difference between the fitted amplitude brightness F1 and the original amplitude brightness value R1, take the absolute value of the difference and perform 11-point sliding average to obtain a sliding average result;

[0108] The sliding average result is judged by using a set threshold, and the first and last pixel numbers of the pixels that are continuously greater than the threshold are extracted.

[0109] Wherein, step (3) is specifically as follows:

[0110] The pixel numbers of the extracted first and last pixels are taken as the independent variables, and their radiance is taken as the dependent variable to perform linear fitting, and the radiance of the pixels between the first and last pixels is fitted to obtain the fitting result.

[0111] Wherein, step (4) is specifically as follows:

[0112] Using a Savitzky-Golay filter to fit the new amplitude brightness value R2 to obtain a fitting radiance F2; or

[0113] The new amplitude brightness value R2 is subjected to sliding fitting by adopting sliding average, Fourier filtering or empirical mode decomposition to obtain the fitting radiance F2.

[0114] Wherein, step (5) is specifically as follows:

[0115] Calculate the difference between the original amplitude brightness value R1 and the fitted radiance F2, and add the background radiance value caused by airglow to obtain the result of removing stray light;

[0116] The background radiance value caused by airglow is determined by MERSI / DNB observations in the Antarctic region during June and July without any spurious nights.

[0117] Optionally, the present application also provides another device for removing stray light from a twilight orbit satellite micro-light channel. Figure 5 As shown, the stray light removal device may include: one or more processors 101, one or more input devices 102, one or more output devices 103 and a memory 104, wherein the processors 101, the input devices 102, the output devices 103 and the memory 104 are interconnected via a bus 105. The memory 104 is used to store a computer program, wherein the computer program includes program instructions, and the processor 101 is configured to call the program instructions to execute the method of the above method embodiment.

[0118] It should be understood that in the embodiment of the present invention, the processor 101 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0119] The input device 102 may include a keyboard, etc., and the output device 103 may include a display (LCD, etc.), a speaker, etc. The memory 104 may include a read-only memory and a random access memory, and provide instructions and data to the processor 101. A portion of the memory 104 may also include a non-volatile random access memory. For example, the memory 104 may also store information about the device type.

[0120] In a specific implementation, the processor 101, input device 102, and output device 103 described in the embodiment of the present invention can execute the implementation method described in the embodiment of the stray light removal method provided by the embodiment of the present invention, which will not be repeated here.

[0121] It should be noted that, regarding the specific implementation of the stray light removal device, please refer to the aforementioned method embodiment part, which will not be repeated here.

[0122] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or substitutions within the technical scope disclosed by the present invention, and these modifications or substitutions should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for removing stray light from a twilight orbit satellite channel, characterized in that: include: Get MERSI / LLB original images; The original image includes only pure nighttime data; Processing the original image using a stray light removal algorithm to obtain a target image; The stray light removal algorithm comprises the following steps: Perform sliding fitting on the original image scan line by scan line, assuming that the amplitude brightness value of a certain scan line is R1, and the fitting radiance is F1; Extract the pixel number and radiance of the first and last pixels of the continuous light pixel according to the difference between the amplitude brightness value R1 and the fitting amplitude brightness F1; Performing linear fitting on the radiance of the continuous light pixel to obtain a fitting result, and using the fitting result to replace the value of the amplitude brightness value R1 of the corresponding pixel to obtain a new amplitude brightness value R2; Perform sliding fitting on the new amplitude brightness value R2 to obtain fitting radiance F2; According to the amplitude brightness value R1, the fitting radiance F2 and the preset background radiance value, the target image after removing the stray light is obtained.

2. The method for removing stray light from the twilight orbit satellite micro-light channel according to claim 1, characterized in that: Before acquiring the MERSI / LLB original image, the method further includes: The solar zenith angle data of MERSI / LLB is read, and pure night data is selected therefrom; the pure night data is data with a solar zenith angle greater than 101°.

3. The method for removing stray light from the twilight orbit satellite micro-light channel as claimed in claim 2, characterized in that: The amplitude brightness value R1 is fitted to obtain the fitted radiance F1, which is specifically: Use Savitzky-Golay filter, take 61 pixels as window, and use cubic polynomial to make sliding fit to the amplitude brightness value R1 to get the fitting radiance F1; or The amplitude brightness value R1 is subjected to sliding fitting by sliding average, Fourier filtering or empirical mode decomposition to obtain the fitting radiance F1.

4. The method for removing stray light from the twilight orbit satellite micro-light channel as claimed in claim 2, characterized in that: The pixel number and radiance of the first and last pixels of the continuous light pixel are extracted according to the difference between the amplitude brightness value R1 and the fitting amplitude brightness F1, specifically: Calculate the difference between the fitting amplitude brightness F1 and the amplitude brightness value R1, take the absolute value of the difference and perform 11-point sliding average to obtain a sliding average result; The sliding average result is judged by using a set threshold, pixels that are continuously greater than the threshold are judged as continuous light pixels, and the pixel numbers and radiances of the first and last pixels are extracted.

5. The method for removing stray light from the twilight orbit satellite micro-light channel as claimed in claim 2, characterized in that: The radiance of the continuous light pixel is linearly fitted to obtain the fitting result, which is specifically: The pixel numbers of the extracted first and last pixels are taken as the independent variables, and their radiance is taken as the dependent variable to perform linear fitting, and the radiance of the pixels between the first and last pixels is fitted to obtain the fitting result.

6. The method for removing stray light from the twilight orbit satellite micro-light channel as claimed in claim 2, characterized in that: The new amplitude brightness value R2 is fitted to obtain the fitted radiance F2, specifically: Using a Savitzky-Golay filter to fit the new amplitude brightness value R2 to obtain a fitting radiance F2; or The new amplitude brightness value R2 is subjected to sliding fitting by adopting sliding average, Fourier filtering or empirical mode decomposition to obtain the fitting radiance F2.

7. The method for removing stray light from the twilight orbit satellite micro-light channel as claimed in claim 2, characterized in that: According to the amplitude brightness value R1, the fitting radiance F2 and the preset background radiance value, the target image after removing the stray light is obtained, specifically: Calculate the difference between the amplitude brightness value R1 and the fitted radiance F2, and add the background radiance value caused by airglow to get the result of removing stray light; The background radiance value caused by airglow is determined by MERSI / LLB observations in the Antarctic region during June and July without any spurious nights.

8. A device for removing stray light from a twilight orbit satellite micro-light channel, characterized in that: The method comprises a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program comprises program instructions, and the processor is configured to call the program instructions to execute the method according to any one of claims 1 to 7.

9. A device for removing stray light from a twilight orbit satellite channel, characterized in that: include: An acquisition unit, used for acquiring MERSI / LLB original images; The original image includes only pure nighttime data; A processing unit, configured to process the original image using a stray light removal algorithm to obtain a target image; The stray light removal algorithm comprises the following steps: Perform sliding fitting on the original image scan line by scan line, assuming that the amplitude brightness value of a certain scan line is R1, and the fitting radiance is F1; Extract the pixel number and radiance of the first and last pixels of the continuous light pixel according to the amplitude brightness value R1 and the fitting amplitude brightness F1; Perform linear fitting on the radiance of continuous light pixels to obtain fitting results, and use the fitting results to replace the amplitude brightness value R1 of the corresponding pixel to obtain a new amplitude brightness value R2; Fitting the new amplitude brightness value R2 to obtain a fitting radiance F2; According to the amplitude brightness value R1, the fitting radiance F2 and the preset background radiance value, the target image after removing the stray light is obtained.

Images