Multi-channel thermal infrared surface temperature inversion method and device

Abstract

The invention belongs to the technical field of environment detection, and discloses a multichannel thermal infrared surface temperature inversion method and device, and the method comprises the steps: obtaining a spectrum response function, a water vapor image, a multispectral image, and an infrared image; matching the water vapor image with the infrared image to obtain the atmospheric water vapor content; constructing an analog data set of the atmospheric water vapor content and the channel transmissivity based on the spectrum response function; thermal infrared transmittances of the three thermal infrared channels are solved based on the simulation data set; calculating red light reflectivity and near-infrared reflectivity according to the multispectral image; calculating a normalized vegetation index and vegetation coverage according to the red light reflectivity and the near-infrared reflectivity; determining the surface emissivity based on the normalized vegetation index and the vegetation coverage; inputting the thermal infrared transmissivity and the surface emissivity into a Planck inversion formula; and solving an inverse function of the Planck inversion formula to obtain the surface temperature. According to the invention, the precision of temperature inversion can be improved.

Description

Technical Field

[0001] The present application relates to the field of environmental detection technology, and in particular to a multi-channel thermal infrared surface temperature inversion method and device. Background Art

[0002] So far, the surface temperature inversion algorithms mainly include single window algorithm, split window algorithm, multi-angle algorithm, day / night algorithm, and temperature emissivity separation algorithm. With the development of thermal infrared remote sensing technology, the number of thermal infrared channels of current environmental observation satellites has increased significantly, but the different derivation forms of the split window algorithm only involve two adjacent thermal infrared channels, and most of the known multi-channel algorithms introduce mid-infrared channels on the basis of the split window algorithm to improve the inversion accuracy.

[0003] None of the existing algorithms fully utilize the advantages of multiple thermal infrared channels of satellite remote sensing data. At the same time, in the existing various surface temperature inversion algorithms, the inversion of surface temperature involves the simplification of the Planck function, specifically directly linking radiation and temperature through Taylor expansion to solve the radiation transfer equation set. However, the radiation value corresponding to the Planck function is not strongly linearly related to temperature. Therefore, using temperature to perform Taylor expansion on the Planck function to solve the radiation transfer equation set will introduce certain errors, which will affect the accuracy of the final surface temperature product. Summary of the invention

[0004] The present application provides a multi-channel thermal infrared surface temperature inversion method and device, which can avoid the error introduced when using temperature to perform Taylor expansion of the Planck function to solve the radiation transfer equation set, thereby improving the accuracy of temperature inversion.

[0005] In a first aspect, an embodiment of the present application provides a multi-channel thermal infrared surface temperature inversion method, comprising:

[0006] Obtain the spectral response function of the thermal infrared camera, the water vapor image of the DPC payload, the multispectral images of the red channel and the near-infrared channel, and the infrared images of the three thermal infrared channels;

[0007] Match the water vapor image with the infrared image to obtain the atmospheric water vapor content;

[0008] Construct a simulated data set of atmospheric water vapor content and channel transmittance based on the spectral response function;

[0009] The thermal infrared transmittance of the three thermal infrared channels is obtained based on the simulated data set;

[0010] Calculate red reflectance and near-infrared reflectance based on multispectral images;

[0011] The normalized vegetation index and vegetation coverage were calculated based on the red reflectance and near-infrared reflectance;

[0012] Determine the surface emissivity based on the normalized difference vegetation index and vegetation cover;

[0013] Input each thermal infrared transmittance and surface emissivity into the Planck inversion formula;

[0014] Solve the inverse function of Planck's inversion formula to obtain the surface temperature.

[0015] Furthermore, the method also includes: after obtaining the multispectral image and the infrared image, performing pixel-by-pixel data registration on the multispectral image and the infrared image according to the longitude and latitude information therein.

[0016] Furthermore, the above-mentioned simulation data set of atmospheric water vapor content and channel transmittance constructed based on the spectral response function includes:

[0017] The atmospheric profile data are obtained and input into the atmospheric radiation transfer model to obtain the channel transmittance, atmospheric average upgoing radiation and atmospheric average downgoing radiation, which are convolved with the spectral response function to obtain the simulated data set.

[0018] Furthermore, the above-mentioned calculation of the normalized vegetation index and vegetation coverage based on the red light reflectance and the near-infrared reflectance includes: comparing the difference between the near-infrared reflectance and the red light reflectance with the sum to obtain the normalized vegetation index; comparing the difference between the normalized vegetation index and the bare soil normalized index with the difference between the normalized vegetation index and the bare soil normalized index, squaring the ratio, and obtaining the vegetation coverage.

[0019] Furthermore, the above-mentioned determination of the surface emissivity based on the normalized vegetation index and the vegetation coverage includes:

[0020] The soil surface emissivity was calculated based on the red light reflectance and near infrared reflectance;

[0021] The vegetation surface emissivity was obtained according to the normalized vegetation index;

[0022] The mixed emissivity is calculated based on the soil surface emissivity, vegetation surface emissivity and vegetation coverage;

[0023] The numerical interval of the normalized difference vegetation index is determined, and the soil surface emissivity, the vegetation surface emissivity or the mixed emissivity is selected as the surface emissivity according to the numerical interval.

[0024] Furthermore, the method also includes:

[0025] The atmospheric radiation transfer equation is used to represent the top atmosphere radiation received by each thermal infrared channel;

[0026] Based on the mean value theorem, a simplified equation relationship between atmospheric transmittance, average temperature and top-of-atmosphere radiation is constructed;

[0027] Based on the simplified equation relationship, the radiation transfer equation set of each thermal infrared channel is obtained;

[0028] Taylor expansion is performed on the set of radiation transfer equations to obtain the effective Planck function of each thermal infrared channel;

[0029] By combining the effective Planck functions and eliminating the top atmosphere radiation, we can obtain the Planck inversion formula.

[0030] Furthermore, the above-mentioned input of each thermal infrared transmittance and surface emissivity into the Planck inversion formula includes:

[0031] Calculate the effective wavelength and effective Planck function of each thermal infrared channel based on the spectral response function;

[0032] Each thermal infrared transmittance, surface emissivity, thermal infrared camera radiance, effective wavelength of the second thermal infrared channel and effective Planck function are input into the Planck inversion formula.

[0033] In a second aspect, an embodiment of the present application provides a multi-channel thermal infrared surface temperature inversion device, comprising:

[0034] An acquisition module is used to acquire the spectral response function of the thermal infrared camera, the water vapor image of the DPC payload, the multispectral images of the red channel and the near infrared channel, and the infrared images of the three thermal infrared channels;

[0035] A matching module is used to match the water vapor image with the infrared image to obtain the atmospheric water vapor content;

[0036] A construction module is used to construct a simulated data set of atmospheric water vapor content and channel transmittance based on the spectral response function;

[0037] The transmission solution module is used to solve the thermal infrared transmittance of three thermal infrared channels based on the simulated data set;

[0038] A reflection calculation module is used to calculate red light reflectance and near infrared reflectance based on multispectral images;

[0039] Coverage module, used to calculate the normalized vegetation index and vegetation coverage based on red reflectance and near-infrared reflectance;

[0040] A radiation rate determination module, used for determining the surface radiation rate based on the normalized vegetation index and the vegetation coverage;

[0041] An input module is used to input various thermal infrared transmittances and surface emissivities into the Planck inversion formula;

[0042] The inverse solution module is used to solve the inverse function of Planck's inversion formula to obtain the surface temperature.

[0043] In a third aspect, an embodiment of the present application provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of a multi-channel thermal infrared surface temperature inversion method as in any of the above embodiments are performed.

[0044] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a multi-channel thermal infrared surface temperature inversion method as in any of the above embodiments.

[0045] In summary, compared with the prior art, the technical solution provided in the embodiment of the present application has at least the following beneficial effects:

[0046] The embodiment of the present application provides a multi-channel thermal infrared surface temperature inversion method, which utilizes the strong linear correlation of the radiation relationship between adjacent thermal infrared channels. Therefore, when jointly eliminating the atmospheric influence in the radiation transfer equation set, it is not necessary to convert the radiation value into a linear representation of the brightness temperature, but the radiation value can be directly converted into a linear representation of the radiation value of the adjacent channel to simplify the equation set. And the inverted radiation value finally obtained corresponds to the surface temperature one-to-one according to the inverse transformation of the Planck function, avoiding the error introduced when using the temperature to perform Taylor expansion on the Planck function to solve the radiation transfer equation set, and improving the accuracy of temperature inversion. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A flowchart of a multi-channel thermal infrared surface temperature inversion method provided for an exemplary embodiment of the present application.

[0048] Figure 2 A flow chart of the steps of determining the ground surface emissivity is provided for an exemplary embodiment of the present application.

[0049] Figure 3 A structural diagram of a multi-channel thermal infrared surface temperature inversion device provided as an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0050] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments.

[0051] Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making any creative work shall fall within the scope of protection of this application.

[0052] As the driving force of energy exchange at the surface-atmosphere interface, surface temperature plays an important role in surface-atmosphere interactions at regional and global scales. Accurate surface temperature products can reflect the spatiotemporal dynamic changes of surface radiation under energy balance. Therefore, as a key variable in climatology, environmental science, agriculture and hydrology, it can be used in climate balance, environmental monitoring, land planning, water and heat change, and has been identified as one of the high-priority parameters by the International Geosphere and Biosphere Program (IGBP). In theory, retrieving surface temperature information from thermal infrared satellite remote sensing observation data is an ill-posed problem. That is, if there are X observation channels, then according to the radiation transfer theory, X radiation equations can be obtained, but there are X+1 unknowns (the emissivity of X observation channels and 1 surface temperature). Therefore, no matter how many observation channels are used for simplification, the number of unknowns is always one more than the number of equations. After decades of research, there are many methods that can accurately retrieve surface temperature information from satellite remote sensing observation data under the assumption. So far, the surface temperature inversion algorithms mainly include single window algorithm, split window algorithm, multi-angle algorithm, day / night algorithm, and temperature emissivity separation algorithm. Among them, the single window algorithm requires accurate atmospheric profile and assumes that the surface emissivity information is known; the split window algorithm uses the different atmospheric absorption of two adjacent channels to remove the atmospheric influence, so it does not need atmospheric profile information, but the inversion accuracy is greatly affected by the emissivity accuracy of adjacent channels; the multi-angle algorithm assumes that the surface temperature is independent of VZA (Viewing Zenith Angle) and the atmosphere is horizontally uniform and stable during the observation time, but it requires accurate geometric registration and is only applicable to consistent surfaces; the day / night algorithm does not require surface emissivity information as a priori, but requires accurate atmospheric profile information, and assumes that the surface emissivity does not change much during the day and night. In addition, the day and night image registration has a great impact on the inverted surface temperature, and the overall method is relatively complex; the temperature emissivity separation algorithm can accurately separate the surface temperature and emissivity at the same time, but this method also requires accurate atmospheric correction. Among them, the split window algorithm has become one of the most popular methods in actual operation because it does not require real-time atmospheric profile information. With the development of thermal infrared remote sensing technology, the number of thermal infrared channels of current environmental observation satellites has increased significantly, but the different derivation forms of the split window algorithm only involve two adjacent thermal infrared channels, and most of the known multi-channel algorithms introduce mid-infrared channels on the basis of the split window algorithm to improve the inversion accuracy. It is not difficult to see that the above algorithms have not fully utilized the advantages of multiple thermal infrared channels of satellite remote sensing data. At the same time, in the various surface temperature inversion algorithms mentioned earlier, the inversion of surface temperature involves the simplification of the Planck function, specifically directly linking radiation and temperature through the Taylor expansion, thereby solving the radiation transfer equation set.However, the radiation value corresponding to the Planck function is not strongly linearly correlated with temperature. Therefore, using temperature to perform Taylor expansion on the Planck function to solve the set of radiation transfer equations will introduce certain errors, thereby affecting the accuracy of the final surface temperature product.

[0053] To solve the above mentioned problems, see Figure 1 The embodiment of the present application provides a multi-channel thermal infrared surface temperature inversion method, which may specifically include the following steps:

[0054] Step S1, obtaining the spectral response function of the thermal infrared camera, the water vapor image of the DPC load, the multispectral images of the red light channel and the near infrared channel, and the infrared images of the three thermal infrared channels.

[0055] Among them, the thermal infrared channel cannot be located in the ozone absorption band of 9.4 to 10.0 μm, because if it falls in the ozone absorption band, the brightness temperature observed by the channel will not reflect the information of the surface.

[0056] Specifically, the full-spectrum spectroscopic imager (VIMS) carried by the Earth observation satellite has the technical characteristics of a wide spectral range, high spatial resolution, and high radiation calibration accuracy, covering the spectrum range from visible light, near infrared to thermal infrared bands, with a total of 12 channels. Among them, there are 6 visible light multispectral and short-wave infrared channels with a spatial resolution of 20m; 2 mid-infrared channels with a spatial resolution of 40m; 4 thermal infrared channels with a spatial resolution of 40m, and the central wavelengths are 8.20 (8.01-8.39μm), 8.63 (8.42-8.83μm), 10.80 (10.30-11.30μm) and 11.95 (11.40-12.50μm). This application obtains multispectral images of red light channel 3 and near infrared channel 4, as well as infrared images of thermal infrared channels 10, 11, and 12.

[0057] DPC (Directional Polarization Camera) and VIMS are carried together on earth observation satellites to achieve simultaneous and co-location observations. DPC can provide real-time water vapor content for VIMS's surface temperature inversion.

[0058] Step S2, matching the water vapor image with the infrared image to obtain the atmospheric water vapor content.

[0059] Specifically, the atmospheric water vapor content CWV corresponding to the infrared image can be obtained by data matching between the water vapor image and the infrared image. Since the spatial resolution of the sub-satellite point of the DPC payload is much smaller than that of the thermal infrared image, it can be considered that the thermal infrared image is contained in the water vapor image. Therefore, the latitude and longitude of the first and last pixels of the thermal infrared channel are converted into the corresponding xy coordinates in the water vapor image, thereby confirming the overlapping position of the thermal infrared image in the water vapor image, taking out the overlapping pixels in the water vapor image, and interpolating them to the size of the thermal infrared image, and then the atmospheric water vapor content can be extracted from the interpolated image; the interpolation process can choose to use the bilinear interpolation method. The following is the formula of the bilinear interpolation method:

[0060]

[0061]

[0062]

[0063] Among them, Q 11 , Q 12 , Q 21 , Q 22 represents the four original pixels closest to the interpolation pixel, x1, x2, y1 and y2 are the horizontal and vertical coordinates of the four pixels respectively, and f represents the pixel value corresponding to the pixel. Bilinear interpolation is equivalent to performing two single linear interpolations in the x-axis direction and then using one single linear interpolation in the y-axis direction to calculate the pixel value of the interpolation pixel.

[0064] Step S3, constructing a simulated data set of atmospheric water vapor content and channel transmittance based on the spectral response function.

[0065] Specifically, the atmospheric profile data over land during the day and night obtained in the global atmospheric profile library Seebor V5.0 are first input into the atmospheric radiation transfer model MODTRAN 5.2, and the channel transmittance, atmospheric average upward radiation and atmospheric average downward radiation are solved, and then convolved with the spectral response function to obtain the simulated data set. Among them, the Seebor V5.0 global atmospheric profile library contains 15,704 atmospheric profile data under clear sky conditions, and each atmospheric profile records the temperature, humidity and ozone content of 101 atmospheric pressure layers. These atmospheric profile configuration files come from NOAA-88, ECMWF 60L training set, TIGR-3, 8 NOAA Climate Monitoring and Diagnostic Laboratories (CMDL) ozone sondes and radiosondes, and quality checks are carried out to ensure physical consistency near the top of the troposphere. Physically based features of surface skin temperature and surface emissivity are specified for each atmospheric profile in the data set, providing strong data support for hyperspectral and multispectral atmospheric inversion.

[0066] This application is based on the global atmospheric profile library Seebor V5.0. With the land area > 50% as the condition, the atmospheric profile over land is obtained, and secondary screening is performed through sunrise and sunset time to obtain the atmospheric profile data over land during the day / night.

[0067] Step S4, obtaining the thermal infrared transmittance of the three thermal infrared channels based on the simulated data set.

[0068] After obtaining the simulated data set, the channel transmittance τ can be solved by least squares regression i The relationship between the atmospheric water vapor content w is: i =c i,1 w 2 +c i,2 w+c i,3 , where c i,1 、c i,2 and c i,3 are the regression coefficients obtained by simulation; i is the channel number, and the thermal infrared transmittances of the three thermal infrared channels are expressed as τ 10 , τ 11 , τ 12 .

[0069] Step S5, calculating the red light reflectance and the near infrared reflectance according to the multispectral image.

[0070] Specifically, the formula for calculating the near-infrared reflectivity and red reflectivity of the multispectral image is:

[0071] ρ=π(C+A·DN) / [cosθ s F0]

[0072] DN is the gray value of the multispectral image; A is the gain of the absolute calibration coefficient, C is the intercept of the absolute calibration coefficient, θ s is the solar zenith angle, F0 is related to the VIMS channel, for red channel 3 and near infrared channel 4, the corresponding F 03 =1558.84wm -2 .sr -1 .μm -1 , F 04 =1067.73wm -2 .sr -1 .μm -1 .

[0073] Step S6, calculating the normalized vegetation index and vegetation coverage according to the red light reflectance and the near infrared reflectance.

[0074] Specifically, the difference between the near-infrared reflectance and the red reflectance is compared with the sum to obtain the normalized difference vegetation index NDVI.

[0075] NDVI=(ρ4-ρ3) / (ρ4+ρ3)

[0076] Among them, ρ4 represents the near-infrared reflectivity of near-infrared channel 4, and ρ3 represents the red light reflectivity of red light channel 3.

[0077] Then, the difference between the NDVI and the NDVI is compared with the difference between the NDVI and the NDVI, and the ratio is squared to obtain the vegetation coverage. Calculation formula:

[0078]

[0079] In the formula, FVC represents vegetation coverage, NDVI s Represents the Normalized Difference Index of Bare Soil, NDVI v Represents the Normalized Difference Vegetation Index. Although NDVI v and NDVI s They are not the same across pixels, but they can be approximated as fixed values ​​to maintain spatial consistency. Here, NDVI is chosen. s =0.2, NDVI v =0.86.

[0080] Step S7, determining the surface emissivity based on the normalized vegetation index and the vegetation coverage.

[0081] Step S8: input each thermal infrared transmittance and surface emissivity into the Planck inversion formula.

[0082] Step S9, solving the inverse function of Planck's inversion formula to obtain the surface temperature.

[0083] The multi-channel thermal infrared surface temperature inversion method provided by the above embodiment utilizes the strong linear correlation of the radiation relationship between adjacent thermal infrared channels. Therefore, when jointly eliminating the atmospheric influence in the radiation transfer equation set, it is not necessary to convert the radiation value into a linear representation of the brightness temperature, but the radiation value can be directly converted into a linear representation of the radiation value of the adjacent channel to simplify the equation set. And the inverted radiation value finally obtained corresponds to the surface temperature one-to-one according to the inverse transformation of the Planck function, avoiding the error introduced when using the temperature to perform Taylor expansion on the Planck function to solve the radiation transfer equation set, and improving the accuracy of temperature inversion.

[0084] In some embodiments, the method further includes: after obtaining the multispectral image and the infrared image, performing pixel-by-pixel data registration on the multispectral image and the infrared image according to the longitude and latitude information therein.

[0085] Specifically, since the resolution of the multispectral image is not necessarily the same as that of the infrared image, for example, when the resolution of the multispectral image is 20m, the resolution of the infrared image is 40m, it is necessary to match the data of the two images so that the ground objects corresponding to the pixel data of different images are the same. First, it is necessary to match the data pixel by pixel according to the longitude and latitude information. Since the resolution of the multispectral image is often better than that of the infrared image, that is, the observation range of the multispectral image is narrower for the surface observation range, it is necessary to determine the overlapping range of the corresponding images according to the latitude information of the first pixel of the multispectral image and the infrared image.

[0086] Taking the registration of the red light channel 3 and the thermal infrared channel 10 as an example, specifically, first convert the two images from xy coordinates to longitude and latitude coordinates, and determine the relationship between the latitudes of the first pixels of the red light channel 3 and the thermal infrared channel 10; if the latitude of the first pixel of the red light channel 3 is small, it indicates that only part of the multispectral image is contained in the infrared image, and it is necessary to convert the latitude of the first pixel of the thermal infrared channel 10 and the longitude of the first pixel of the red light channel 3 into the corresponding xy coordinates in the red light channel 3, and thus start the image registration; if the latitude of the first pixel of the red light channel 3 is large, it indicates that the multispectral image is completely contained in the infrared image, and it is necessary to convert the latitude and longitude of the first pixel of the red light channel 3 into the corresponding xy coordinates in the thermal infrared channel 10, and thus start the image registration.

[0087] During the registration process, the image data of red light channel 3 and near infrared channel 4 need to be registered with the data of thermal infrared channels 10, 11, and 12 respectively, that is, a total of six registrations, but because the multispectral image and the infrared image are generated by different cameras of VIMS at the same time and place, the longitude and latitude of the corresponding pixels of red light channel 3 and near infrared channel 4 are the same, and the longitude and latitude of the corresponding pixels of thermal infrared channels 10, 11, and 12 are the same, so they can be processed simultaneously in one registration process. After the image registration, it is also necessary to reduce the resolution of the multispectral image, that is, the pixel average value of the four pixels in the 2*2 range of the multispectral image is used as the new pixel value of the four pixels, so as to match it with the infrared image, that is, the two represent the same ground object target.

[0088] The above embodiment performs pixel-by-pixel registration on the acquired infrared image and multispectral image so that the ground objects corresponding to the pixel data of different images are the same, thereby improving the accuracy of inverting the surface temperature by connecting multiple channels.

[0089] See also Figure 2 In one embodiment, the above-mentioned determination of the surface emissivity based on the normalized vegetation index and the vegetation coverage includes:

[0090] Step S71, calculating the soil surface emissivity based on the red light reflectivity and the near infrared reflectivity.

[0091] Step S72, obtaining vegetation surface emissivity according to the normalized vegetation index.

[0092] Through the simulated data set constructed above, the reflectance of the red light channel and the near-infrared channel are calculated according to the effective emissivity and radiation transfer equation of the surface sample, so that the empirical statistical relationship between the emissivity and reflectivity of the visible light and near-infrared channels can be solved. The empirical statistical relationship between the emissivity and the normalized difference vegetation index NDVI is:

[0093]

[0094] ε vi =b i0 +b i1 NDVI

[0095] where ε si and ε vi Respectively represent the soil surface emissivity and vegetation surface emissivity (complete vegetation coverage), n = 2, ρ j The reflectivity of the red band includes the red light reflectivity ρ1 (ρ1 is equal to ρ3 in the above embodiment) and the near infrared reflectivity ρ2 (ρ2 is equal to ρ4 in the above embodiment); a i0 、a ij , b i0 and b i1 is the regression coefficient.

[0096] Step S73, calculating the mixed emissivity according to the soil surface emissivity, the vegetation surface emissivity and the vegetation coverage.

[0097] For soil / vegetation mixed pixels, the mixed emissivity ε is calculated based on the soil surface emissivity and the vegetation surface emissivity. i :

[0098] ε i =ε vi f v +ε si (1-f v )+C i

[0099] C i =(1-ε si )(1-f v )Fε vi

[0100] Among them, f v is the vegetation coverage, C i Represents the surface emissivity increment caused by the cavity effect caused by multiple scattering within the pixel; F is a model factor in the range of 0-1 and is set to an average value of 0.55.

[0101] Step S74, determining the numerical interval of the normalized vegetation index, and selecting the soil surface emissivity, the vegetation surface emissivity or the mixed emissivity as the surface emissivity according to the numerical interval.

[0102] Specifically, the surface specific emissivity is obtained based on the normalized vegetation index:

[0103]

[0104] It can be seen from the above formula that in the specific implementation process, when the numerical interval is less than 0.2, the soil surface emissivity is used as the surface emissivity; when the numerical interval is greater than 0.86, the vegetation surface emissivity is used as the surface emissivity; when the numerical interval is 0.2-0.86, the mixed emissivity is used as the surface emissivity.

[0105] The above embodiment determines the appropriate surface emissivity based on the numerical range of the normalized difference vegetation index, reduces the surface temperature calculation error caused by different surface environments, and improves the calculation accuracy of the surface temperature.

[0106] In some embodiments, the method further comprises:

[0107] The atmospheric radiation transfer equation is used to represent the top atmosphere radiation received by each thermal infrared channel;

[0108] Based on the mean value theorem, a simplified equation relationship between atmospheric transmittance, average temperature and top-of-atmosphere radiation is constructed;

[0109] Based on the simplified equation relationship, the radiation transfer equation set of each thermal infrared channel is obtained;

[0110] Taylor expansion is performed on the set of radiation transfer equations to obtain the effective Planck function of each thermal infrared channel;

[0111] By combining the effective Planck functions and eliminating the top atmosphere radiation, we can obtain the Planck inversion formula.

[0112] In some embodiments, the above-mentioned inputting each thermal infrared transmittance and surface emissivity into the Planck inversion formula includes:

[0113] Step S81, calculating the effective wavelength and effective Planck function of each thermal infrared channel based on the spectral response function.

[0114] The formula for calculating the effective wavelength, effective Planck function and effective emissivity based on the spectral response function is:

[0115]

[0116] Where C(λ) represents the calculation parameters, namely wavelength λ, Planck function B(λ, T) and emissivity spectrum ε(λ); C effIndicates the effective value of the corresponding calculation parameter, that is, the effective wavelength λ eff 、Effective Planck function B(λ eff , T) and effective emissivity ε eff When the calculation parameter is the wavelength of a channel, the corresponding calculated value is the effective wavelength of the channel; f(λ) represents the spectral response function.

[0117] Specifically, when constructing the empirical statistical relationship between emissivity and reflectivity, the formula here can be used to calculate the effective emissivity of the red light channel and the near-infrared channel, and the known radiation transfer equation can be combined to solve the empirical statistical relationship between the channel emissivity and reflectivity of each red band, thereby using the empirical statistical relationship to calculate the true soil surface emissivity in step S71.

[0118] Step S82, inputting each thermal infrared transmittance, surface emissivity, thermal infrared camera radiance, effective wavelength of the second thermal infrared channel and effective Planck function into the Planck inversion formula.

[0119] According to the determined surface emissivity, thermal infrared camera radiance and thermal infrared transmittance obtained based on the atmospheric water vapor content of the DPC payload, the thermal infrared channels 10, 11 and 12 of VIMS are selected as the target channels. The emission radiation of the channel corresponding to the surface temperature is inverted according to the following formula. The subscripts 10, 11 and 12 of the following parameters represent the corresponding thermal infrared channels:

[0120] B 11 (T S )=XL 10 -YL 11 -ZL 12 -Q

[0121]

[0122] In the formula, ε 10 , ε 11 , ε 12 and τ 10 , τ 11 , τ 12 Respectively represent the surface emissivity and thermal infrared transmittance corresponding to thermal infrared channels 10, 11, and 12; A i , C i , D i represents the combined parameter related to channel i, surface emissivity and atmospheric transmittance; L 10 , L 11 , L 12 B is the radiation received by thermal infrared channels 10, 11, and 12; 11 represents the effective Planck function of thermal infrared channel 11; k 10 , k 12It is a constant determined by the radiometry of thermal infrared channels 10 and 11, 12 and 11.

[0123] Finally, the surface temperature is solved according to the Planck inverse function:

[0124]

[0125] C1=2hc 2 =1.19104356*10 -6 Wm -2

[0126]

[0127] λ 11 =1O.80um

[0128] In some embodiments, the method further comprises:

[0129] Step S01, using the atmospheric radiation transfer equation to represent the top atmosphere radiation received by each thermal infrared channel.

[0130] Specifically, under clear and cloudless weather conditions, according to the atmospheric radiation transfer equation, the thermal radiation L of the top atmosphere received by the satellite sensor thermal infrared channel i is i It can be expressed as:

[0131]

[0132] Step S02: construct a simplified equation relationship between the atmospheric transmittance, average temperature and top atmosphere radiation based on the mean value theorem.

[0133] Step S03, obtaining a set of radiation transmission equations for each thermal infrared channel according to the simplified equation relationship.

[0134] The upward atmospheric radiation is converted into and downwelling atmospheric radiation The radiation transfer equation can be simplified to the following form by using atmospheric transmittance, upward average temperature and downward average temperature:

[0135]

[0136]

[0137] A i =ε i τ i , C i =(1-ε i )(1-τ i )τ i , D i =1-τ i

[0138] Among them, L i represents the top atmosphere radiance received by the sensor of channel i, ε i is the firing rate of channel i, τ i is the transmittance of channel i, and is the average temperature of the atmosphere going down and the average temperature of the atmosphere going up, B i is the effective Planck function of channel i, expressed as the integral of the Planck function and the spectral response function f(λ):

[0139]

[0140] Therefore, for adjacent thermal infrared channels i, s and t, the set of radiation transfer equations can be obtained:

[0141]

[0142] Step S04: Taylor expansion is performed on the radiation transfer equation set to obtain the effective Planck function of each thermal infrared channel.

[0143] Step S05, combining the effective Planck functions, eliminating the top atmosphere radiation, and obtaining the Planck inversion formula.

[0144] Specifically, considering that the Planck radiation relationship between adjacent channels shows a strong linear correlation, a Taylor expansion can be performed on it, that is, the Planck radiation relationship between adjacent channels can be expressed as a linear relationship:

[0145]

[0146]

[0147]

[0148] in, K i，2 (K s，2 )=c1 / λ i are all constants.

[0149] Since the thermal infrared band is between 8.0 and 14.0 μm, which belongs to the long-wave range, the downward atmospheric radiation is usually slightly greater than the upward atmospheric radiation, and Therefore, there is a widespread B s (T S ), and All located in L s In a small radiation range centered at L s Expanding it, we get the following relationship:

[0150]

[0151] Therefore, the radiation transfer equation of channel i can be linearly expressed by the effective Planck function of channel s:

[0152]

[0153] Similarly, the radiation transfer equation of channel t is linearly expressed by the effective Planck function of channel s, and finally the radiation transfer equation set of adjacent thermal infrared channels i, s and t can be simplified to the following form:

[0154]

[0155] Combined elimination of atmospheric radiation and The influence of B s (T S ) is expressed in the form of adjacent channel radiance:

[0156] B s (T S )=XL i -YL s -ZL t -Q

[0157]

[0158] Substitute the effective wavelength of the channel center into the Planck equation for inverse transformation and simplification, and construct the surface temperature calculation equation to calculate the surface temperature:

[0159]

[0160] C1=2hc 2 =1.19104356*10 -6 Wm -2

[0161]

[0162] Among them, T S is the surface temperature; B s (T S ) is the surface temperature T of channel s S Blackbody radiation; λ s is the effective wavelength of channel s.

[0163] The above-mentioned embodiment models the radiation relationship between multiple adjacent channels in the thermal infrared band, and from the perspective of the radiation transfer equation, after rigorous mathematical derivation, obtains the expression for surface temperature inversion. Compared with the previous split-window algorithm and multi-channel algorithm, this application constructs the relationship between the radiation value corresponding to the surface temperature and the radiation value of the adjacent thermal infrared channel, rather than the empirical relationship between the surface temperature and the brightness temperature of the adjacent thermal infrared channel. Unlike the simple extension of the split-window algorithm, the surface temperature inversion expression constructed by this application links the three thermal infrared channels to each other, and the resulting surface temperature product is more accurate and more robust.

[0164] See also Figure 3 Another embodiment of the present application provides a multi-channel thermal infrared surface temperature inversion device, comprising:

[0165] The acquisition module 101 is used to acquire the spectral response function of the thermal infrared camera, the water vapor image of the DPC load, the multispectral images of the red light channel and the near infrared channel, and the infrared images of the three thermal infrared channels.

[0166] The matching module 102 is used to match the water vapor image with the infrared image to obtain the atmospheric water vapor content.

[0167] The construction module 103 is used to construct a simulation data set of atmospheric water vapor content and channel transmittance based on the spectral response function.

[0168] The transmission solution module 104 is used to obtain the thermal infrared transmittance of the three thermal infrared channels based on the simulation data set.

[0169] The reflection calculation module 105 is used to calculate the red light reflectance and the near infrared reflectance according to the multi-spectral image.

[0170] The coverage module 106 is used to calculate the normalized vegetation index and vegetation coverage according to the red light reflectance and the near infrared reflectance.

[0171] The radiance determination module 107 is used to determine the surface radiance based on the normalized vegetation index and the vegetation coverage.

[0172] The input module 108 is used to input each thermal infrared transmittance and surface emissivity into the Planck inversion formula.

[0173] The inverse solution module 109 is used to solve the inverse function of the Planck inversion formula to obtain the surface temperature.

[0174] In some embodiments, the device further includes: an image registration module for performing pixel-by-pixel data registration on the multispectral image and the infrared image according to the latitude and longitude information therein after obtaining the multispectral image and the infrared image.

[0175] The specific limitations of a multi-channel thermal infrared surface temperature inversion device provided in this embodiment can be found in the above embodiment of a multi-channel thermal infrared surface temperature inversion method, which will not be repeated here. Each module in the above-mentioned multi-channel thermal infrared surface temperature inversion device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0176] The embodiment of the present application provides a computer device, which may include a processor, a memory, a network interface and a database connected through a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the processor executes the steps of a multi-channel thermal infrared surface temperature inversion method as in any of the above embodiments.

[0177] The working process, working details and technical effects of the computer device provided in this embodiment can be found in the above embodiment of a multi-channel thermal infrared surface temperature inversion method, which will not be described in detail here.

[0178] An embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of a multi-channel thermal infrared surface temperature inversion method as in any of the above embodiments are implemented. Wherein, the computer-readable storage medium refers to a carrier for storing data, which may include but is not limited to a floppy disk, an optical disk, a hard disk, a flash memory, a USB flash drive and / or a memory stick (Memory Stick), etc., and the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The working process, working details and technical effects of the computer-readable storage medium provided in this embodiment can be found in the above embodiment of a multi-channel thermal infrared surface temperature inversion method, and will not be repeated here.

[0179] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0180] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification. The above-mentioned embodiments only express several implementation methods of the present application, and their descriptions are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that for ordinary technicians in this field, without departing from the concept of the present application, several deformations and improvements can be made, which all belong to the protection scope of the present application. Therefore, the protection scope of the patent of this application shall be based on the attached claims.

Claims

1. A multi-channel thermal infrared surface temperature inversion method, characterized in that: include: Obtain the spectral response function of the thermal infrared camera, the water vapor image of the DPC payload, the multispectral images of the red channel and the near-infrared channel, and the infrared images of the three thermal infrared channels; Matching the water vapor image with the infrared image to obtain the atmospheric water vapor content; Constructing a simulation data set of the atmospheric water vapor content and channel transmittance based on the spectral response function; Obtaining the thermal infrared transmittance of the three thermal infrared channels based on the simulation data set; Calculating red light reflectance and near infrared reflectance according to the multispectral image; The normalized vegetation index and vegetation coverage were calculated based on the red reflectance and near-infrared reflectance; Determining a surface emissivity based on the normalized difference vegetation index and the vegetation coverage; Inputting each of the thermal infrared transmittance and the surface emissivity into the Planck inversion formula; The inverse function of the Planck inversion formula is solved to obtain the surface temperature.

2. The multi-channel thermal infrared surface temperature inversion method according to claim 1 is characterized in that: Also includes: After the multispectral image and the infrared image are obtained, pixel-by-pixel data registration is performed on the multispectral image and the infrared image according to the longitude and latitude information therein.

3. The multi-channel thermal infrared surface temperature inversion method according to claim 1 is characterized in that: The step of constructing the simulated data set of the atmospheric water vapor content and channel transmittance based on the spectral response function includes: Atmospheric profile data are obtained and input into an atmospheric radiation transfer model, channel transmittance, atmospheric average up-radiation and atmospheric average down-radiation are obtained by solving the model, and the data are convolved with the spectral response function to obtain the simulated data set.

4. The multi-channel thermal infrared surface temperature inversion method according to claim 1 is characterized in that: The normalized vegetation index and vegetation coverage are calculated based on the red light reflectance and the near infrared reflectance, including: Comparing the difference and sum of the near-infrared reflectivity and the red reflectivity to obtain the normalized vegetation index; The difference between the normalized vegetation index and the bare soil normalized index is compared with the difference between the normalized vegetation index and the bare soil normalized index, and the ratio is squared to obtain the vegetation coverage.

5. The multi-channel thermal infrared surface temperature inversion method according to claim 1 is characterized in that: The determining of the surface emissivity based on the normalized vegetation index and the vegetation coverage comprises: Calculate the soil surface emissivity according to the red light reflectivity and the near infrared reflectivity; Obtaining vegetation surface emissivity according to the normalized vegetation index; Calculating mixed emissivity according to the soil surface emissivity, the vegetation surface emissivity and the vegetation coverage; The numerical interval of the normalized vegetation index is determined, and the soil surface emissivity, the vegetation surface emissivity or the mixed emissivity is selected as the surface emissivity according to the numerical interval.

6. The multi-channel thermal infrared surface temperature inversion method according to claim 1, characterized in that: Also includes: The atmospheric radiation transfer equation is used to represent the top atmosphere radiation received by each thermal infrared channel; Based on the mean value theorem, a simplified equation relationship between atmospheric transmittance, average temperature and top-of-atmosphere radiation is constructed; According to the simplified equation relationship, a set of radiation transfer equations of each thermal infrared channel is obtained; Performing Taylor expansion on the radiation transfer equation set to obtain the effective Planck function of each thermal infrared channel; The effective Planck functions are combined to eliminate the top atmosphere radiation, thereby obtaining the Planck inversion formula.

7. The multi-channel thermal infrared surface temperature inversion method according to claim 5, characterized in that: The step of inputting the thermal infrared transmittance and the surface emissivity into the Planck inversion formula comprises: Calculate the effective wavelength and effective Planck function of each thermal infrared channel based on the spectral response function; The thermal infrared transmittance, the surface emissivity, the thermal infrared camera radiance, the effective wavelength of the second thermal infrared channel and the effective Planck function are input into the Planck inversion formula.

8. A multi-channel thermal infrared surface temperature inversion device, characterized in that: include: An acquisition module is used to acquire the spectral response function of the thermal infrared camera, the water vapor image of the DPC payload, the multispectral images of the red channel and the near infrared channel, and the infrared images of the three thermal infrared channels; A matching module, used for matching the water vapor image with the infrared image to obtain the atmospheric water vapor content; A construction module, used to construct a simulation data set of the atmospheric water vapor content and channel transmittance based on the spectral response function; A transmission solution module, used for solving the thermal infrared transmittance of the three thermal infrared channels based on the simulation data set; A reflection calculation module, used for calculating red light reflectance and near infrared reflectance according to the multispectral image; Coverage module, used to calculate the normalized vegetation index and vegetation coverage based on red reflectance and near-infrared reflectance; A radiation rate determination module, used for determining the surface radiation rate based on the normalized vegetation index and the vegetation coverage; An input module, used for inputting each of the thermal infrared transmittance and the surface emissivity into a Planck inversion formula; The inverse solution module is used to solve the inverse function of the Planck inversion formula to obtain the surface temperature.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the multi-channel thermal infrared surface temperature inversion method as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the multi-channel thermal infrared surface temperature inversion method as claimed in any one of claims 1 to 7 are implemented.

Images