A method and system for correcting changes in infrared radiation observed by a remote sensing satellite from the earth's surface

Abstract

The application discloses a remote sensing satellite observation ground surface infrared radiation change correction method and system, relates to the infrared remote sensing monitoring field, and comprises the following steps: acquiring multi-scene infrared band data and corresponding sensor observation zenith angles, performing radiation calibration and atmospheric correction on the infrared band data, and obtaining ground surface radiance or ground surface temperature data; selecting pseudo-invariant feature points in a target region based on the ground surface radiance or ground surface temperature data, and establishing a relationship model between the sensor observation zenith angles and the ground surface radiation values based on the pseudo-invariant feature points; setting a standard reference observation zenith angle, and calculating normalized observation radiation values under different observation zenith angles to corrected radiation values under the standard reference observation zenith angle based on the relationship model; applying the normalized correction coefficient to the whole scene image, and performing multiplication correction on the original radiation values of each pixel; and repeatedly performing the above steps on multi-temporal infrared remote sensing images to generate an angle-normalized time series radiation data set.

Description

Technical Field

[0001] This invention relates to the field of infrared remote sensing monitoring, specifically to a method and system for correcting changes in infrared radiation on the Earth's surface observed by remote sensing satellites. Background Technology

[0002] Infrared remote sensing is a crucial tool for monitoring surface temperature, energy balance, fires, and environmental changes. Due to the need for satellite orbit, attitude adjustments, or side-swing observations, observations of the same ground target are often conducted at different times and under different sensor deflection angles. Changes in the sensor's observation angle alter the optical path, atmospheric effects, and the effective radiation direction of the target object, leading to significant differences in the apparent radiance of the same object observed at different angles (i.e., the angle effect). This radiation difference, which is not a reflection of the actual changes in the ground object itself, severely interferes with the application of time-series-based infrared remote sensing data, such as long-term trend analysis of surface temperature and dynamic monitoring of disasters.

[0003] Currently, traditional radiometric calibration and atmospheric correction can only eliminate the sensor's own response and atmospheric effects along the vertical path, but cannot effectively correct radiometric distortion caused by changes in the observation angle. Some existing angle correction models often rely on complex two-way reflection / radiation distribution function (BRDF / BTDF) models, which require multi-angle synchronous observation data or prior knowledge base support. They are computationally complex, have poor universality, and are difficult to apply operationally to historical archive data or satellite observation sequences from a single angle.

[0004] Therefore, there is an urgent need for a method that does not rely on multi-angle synchronous observation, is computationally efficient, and can effectively normalize the observed radiation values ​​under different deflection angles, in order to obtain time-series infrared radiation data that truly reflects changes in the physical state of the Earth's surface. Summary of the Invention

[0005] To address the aforementioned shortcomings in existing technologies, this invention provides a method and system for correcting changes in infrared radiation on the Earth's surface observed by remote sensing satellites, which solves the problem of the difficulty in eliminating the radiation angle effect caused by changes in the observation angle, a common issue in infrared remote sensing time series applications.

[0006] To achieve the aforementioned objectives, the present invention employs the following technical solution: a method for correcting changes in infrared radiation of the Earth's surface observed by remote sensing satellites, comprising the following steps: S1: Acquire multi-scene infrared band data and corresponding sensor observation zenith angles, and perform radiometric calibration and atmospheric correction on the infrared band data to obtain surface radiance or surface temperature data. S2: Based on surface radiance or surface temperature data, select pseudo-invariant feature points within the target area, and establish a model of the relationship between the sensor-observed zenith angle and the surface radiation value based on the pseudo-invariant feature points; S3: Set the standard reference observation zenith angle, and calculate the normalized radiance values ​​of the observed zenith angles to the standard reference observation zenith angle based on the relational model; S4: Apply the normalized correction coefficient to the entire scene image and perform multiplicative correction on the original radiometric value of each pixel; S5: Repeat steps S1-S4 on multi-temporal infrared remote sensing images to generate angle-normalized time-series radiation datasets, thereby realizing the correction of changes in surface infrared radiation observed by remote sensing satellites.

[0007] The beneficial effects of the above scheme are: the present invention takes multi-temporal infrared observation data as input, and achieves unified normalization processing of infrared radiation values ​​under different observation zenith angle conditions without relying on multi-angle synchronous observation and complex prior radiation distribution models, thereby obtaining stable time series infrared remote sensing products that truly reflect the changes in the physical state of the Earth's surface.

[0008] Furthermore, the acquisition of multiple infrared band data and corresponding sensor-observed zenith angles, and the radiometric calibration and atmospheric correction of the infrared band data to obtain surface radiance or surface temperature data, includes: Multiple infrared band data from the same remote sensing satellite platform observed the same target area at different times are acquired. Radiometric calibration and atmospheric correction are performed on each data scene to obtain surface radiance or surface temperature data. At the same time, the sensor's zenith angle for each pixel in each data scene is extracted from the satellite auxiliary data.

[0009] The beneficial effects of the above-mentioned further scheme are: to acquire multi-temporal satellite infrared remote sensing data and corresponding sensor-based Earth observation zenith angle data, and to preprocess the data to obtain surface radiation parameter data, namely surface radiance or surface temperature.

[0010] Furthermore, the relationship model between the zenith angle observed by the sensor and the surface radiation value is as follows:

[0011] in, This represents the surface radiation value. For relational models, To enable sensors to observe the zenith angle from the ground, This is the residual.

[0012] The beneficial effect of the above-mentioned further scheme is that, based on the above calculation formula, a relationship model between the sensor-observed zenith angle and the surface radiation value is established.

[0013] Furthermore, the relational model The construction methods include: Cosine model based on the Lambertian approximation:

[0014] in, This is the target radiation signal received by the sensor during vertical observation, and serves as the model's baseline reference value; A correction model based on atmospheric radiative transfer: a relational model is generated through simulation using simplified atmospheric radiative transfer equations. ; Statistical Nonparametric Fitting Model: Utilizing Pseudo-Invariant Feature Points The L-shaped scatter plot was used to obtain the relational model through nonparametric fitting using local weighted regression or spline functions. .

[0015] The beneficial effect of the above-mentioned further solutions is that the present invention can construct a relational model based on any of the above methods. .

[0016] Furthermore, the method for selecting pseudo-invariant feature points within the target area includes: using the stability of vegetation index, surface reflectance, or nighttime light data over a long time series for screening, and selecting several uniform pixels whose surface physical properties remain stable during the observation period as pseudo-invariant feature points.

[0017] The beneficial effect of the above further scheme is that, within the target area, several uniform pixels whose surface physical properties remain stable during the observation period are selected as pseudo-invariant feature points (PIFs). Based on these PIFs, the sensor-observed zenith angle is established. A model relating the observed surface radiation value L to the observed surface radiation value.

[0018] Furthermore, the formula for calculating the corrected radiation value is as follows:

[0019] in, To correct the radiation value, For arbitrary observation of zenith angle The observed radiation values ​​below, This is a model relating the standard reference observation zenith angle to the Earth's surface radiation value. For arbitrary observation of zenith angle A model relating to surface radiation values. For the standard reference observation zenith angle, For arbitrary observation of zenith angle The normalized correction coefficient.

[0020] The beneficial effect of the above further scheme is that the correction process is achieved through the above formula. That is, for the angle Normalized correction coefficient This is used for subsequent multiplication correction of the original radiation value of each pixel.

[0021] Furthermore, the normalized correction coefficient is applied to the entire scene image, and the original radiometric value of each pixel is multiplied for correction. The calculation formula is as follows:

[0022] in, The corrected radiation value, The original radiation value. This is the normalized correction coefficient.

[0023] The beneficial effect of the above further scheme is that, for each pixel in the image, based on its corresponding observed zenith angle... Using the continuous correction coefficient field obtained by its PIF category or spatial interpolation, the original radiance value is multiplicatively corrected to obtain all pixels normalized to the standard reference angle. Radiation data below.

[0024] The technical solution adopted in this invention is: a remote sensing satellite observation surface infrared radiation change correction system, comprising: Data preprocessing module: used to acquire multiple infrared band data and corresponding sensor observation zenith angles, and to perform radiometric calibration and atmospheric correction on the infrared band data to obtain surface radiance or surface temperature data. Model building module: Used to select pseudo-invariant feature points within the target area based on surface radiance or surface temperature data, and to establish a model of the relationship between sensor-observed zenith angle and surface radiation value based on the pseudo-invariant feature points; Corrected radiation calculation module: used to set the standard reference observation zenith angle and calculate the normalized corrected radiation value of the observed radiation value under different observation zenith angles to the standard reference observation zenith angle based on the relational model; Global Correction Module: This module applies normalized correction coefficients to the entire scene image, performing multiplicative correction on the original radiometric value of each pixel. Product generation module: This module is used to repeatedly perform the above processing on multi-temporal infrared remote sensing images to generate angle-normalized time-series radiation datasets, thereby realizing the correction of changes in infrared radiation on the Earth's surface observed by remote sensing satellites.

[0025] The beneficial effects of the above scheme are as follows: This scheme provides a remote sensing satellite observation surface infrared radiation change correction system. By treating the systematic radiation distortion caused by the observation angle as a modelable and normalizable observation geometric effect, a "observation zenith angle-radiation response" relationship model is constructed through pseudo-invariant feature points. The angle normalization correction coefficient is calculated and applied to the whole scene image to achieve a unified benchmark of radiation values ​​in the angular dimension. Attached Figure Description

[0026] Figure 1 This is a flowchart of a method for correcting changes in infrared radiation on the Earth's surface observed by remote sensing satellites.

[0027] Figure 2 This is a schematic diagram of an experiment observing small angle changes.

[0028] Figure 3 This is a schematic diagram of an experiment observing large-angle changes.

[0029] Figure 4 Actual zenith angle Geometric diagram of positional deviation.

[0030] Figure 5 This is a real-world image of the sample arrangement observed in a multi-grain size combination.

[0031] Figure 6 The image shows the results of the sandstone mineral composition identification experiment.

[0032] Figure 7 for Extraction diagram.

[0033] Figure 8 This is a graph showing the infrared radiation temperature variation characteristics of marble samples at four survey line azimuths obtained from observations at different zenith angles.

[0034] Figure 9 The image shows the infrared radiation temperature variation characteristics of granite samples at four survey line azimuths obtained from observations at different zenith angles.

[0035] Figure 10 The image shows the infrared radiation temperature variation characteristics of sandstone samples at four survey line azimuths obtained from observations at different zenith angles.

[0036] Figure 11 For different lithologies and different observation zenith angles T Characteristic diagram of particle size variation. Detailed Implementation

[0037] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0038] Example 1, such as Figure 1 As shown, a method for correcting changes in infrared radiation of the Earth's surface observed by remote sensing satellites includes the following steps: S1: Acquire multi-scene infrared band data and corresponding sensor observation zenith angles, and perform radiometric calibration and atmospheric correction on the infrared band data to obtain surface radiance or surface temperature data, including: Multiple infrared band data from the same remote sensing satellite platform observed the same target area at different times are acquired. Radiometric calibration and atmospheric correction are performed on each data scene to obtain surface radiance or surface temperature data. At the same time, the sensor's zenith angle for each pixel in each data scene is extracted from the satellite auxiliary data.

[0039] S2: Based on surface radiance or surface temperature data, select pseudo-invariant feature points within the target area, and establish a model of the relationship between the sensor-observed zenith angle and the surface radiation value based on the pseudo-invariant feature points; The model relating the zenith angle observed by the sensor to the Earth's surface radiation value is as follows:

[0040] in, This represents the surface radiation value. For relational models, To enable sensors to observe the zenith angle from the ground, This is the residual.

[0041] relational model It can be preferably an analytical function based on physical mechanisms or an empirical fitting function based on statistical learning.

[0042] The relation model The construction methods include: Cosine model based on the Lambertian approximation:

[0043] in, For the sensor to observe vertically (observing the zenith angle) The target radiation signal received at 0° is the model's baseline reference value.

[0044] A correction model based on atmospheric radiative transfer: a relational model is generated through simulation using simplified atmospheric radiative transfer equations. ; Statistical Nonparametric Fitting Model: Utilizing Pseudo-Invariant Feature Points The L-shaped scatter plot was used to obtain the relational model through nonparametric fitting using local weighted regression or spline functions. .

[0045] Methods for selecting pseudo-invariant feature points within the target area include: using the stability of long-term vegetation index, surface reflectance, or nighttime light data for screening, and selecting several uniform pixels whose surface physical properties remain stable during the observation period as pseudo-invariant feature points.

[0046] S3: Set the standard reference observation zenith angle, and calculate the normalized radiance values ​​of the observed zenith angles to the standard reference observation zenith angle based on the relational model; In remote sensing satellite observations of geothermal infrared radiation, the infrared brightness temperature of the same surface target often varies significantly under different observation deflection angles due to the influence of orbital scanning and sensor deflection imaging. This difference mainly stems from directional radiation distortion caused by changes in observation geometry, rather than actual changes in the surface thermal state, easily leading to false information in multi-temporal data. To address this issue, focusing on "lens deflection angle normalization," a directional normalization coefficient corresponding to the observation deflection angle is calculated to uniformly normalize and correct the brightness temperature at different angles. This eliminates systematic errors caused by changes in remote sensing satellite observation angles, improves the consistency and reliability of multi-temporal infrared remote sensing data, and provides stable angle standardization support for accurate monitoring of surface radiation changes.

[0047] In thermal infrared imaging, the failure to consider variations in zenith angle can lead to differences in radiation transmission paths, thus affecting the received radiation brightness and temperature. To eliminate this effect, zenith angle normalization is employed to normalize radiation data from different observation angles, ensuring their comparability.

[0048]

[0049] The above formula preliminarily establishes the infrared radiation temperature. Physical temperature T d Target rock emissivity Emissivity of the surrounding environment of the object The quantitative relationship was established using infrared brightness temperature observations of rock fragments. The rock fragment emissivity calculated from its physical temperature includes the true emissivity of the target rock fragments. and environmental radiative emissivity The combined effects of these factors are termed rock cutting pseudo-emissivity in this invention, and are related to the experimental observation environment, namely:

[0050] In actual space-to-ground observations, the influence of changes in the observed zenith angle also exists, and zenith angle correction is crucial for the inversion calculation of ground temperature. Considering the zenith angle... The impact of change, formula = It can be transformed into:

[0051] in, This represents the gray body radiation energy measured by an infrared thermal imager. This represents the lower limit of the observation band for infrared thermal imagers. This represents the upper limit of the observation band for infrared thermal imagers. This refers to the energy proportion coefficient related to the frequency band. This is the Stefan-Boltzmann constant. Let be the value of the i-th specific zenith angle.

[0052] The normalization of the observed zenith angle effectively solved the problem of the influence of zenith angle variations on radiation temperature. To mitigate the impact of the target's false emissivity and eliminate observational bias, refer to the solution of multi-angle LST.

[0053] Pair Make the following changes:

[0054] In the above formula, The measured infrared brightness temperature of the rock cuttings is given in K. The surface emissivity of the target rock fragments; Let K be the surface physical temperature of the target rock fragment. The emissivity of the environment surrounding the observed object. This refers to the change in zenith angle; The directional index of emissivity change caused by zenith angle variation; for The calculated emissivity value obtained without considering the angle effect; for The physical temperature actually measured at the location.

[0055] The directionality of the effect of zenith angle on emissivity can be expressed by the normalized directivity index. To depict, It is expressed as follows:

[0056] The above formula, The minimum emissivity of the target ground feature without considering its orientation; The maximum emissivity of the target ground object without considering its orientation.

[0057] The formula for calculating the corrected radiation value is:

[0058] in, To correct the radiation value, For arbitrary observation of zenith angle The observed radiation values ​​below, This is a model relating the standard reference observation zenith angle to the Earth's surface radiation value. For arbitrary observation of zenith angle A model relating to surface radiation values. For the standard reference observation zenith angle, For arbitrary observation of zenith angle The normalized correction coefficient.

[0059] S4: Apply the normalized correction coefficient to the entire image, and perform multiplicative correction on the original radiometric value of each pixel. The calculation formula is as follows:

[0060] in, The corrected radiation value, The original radiation value. This is the normalized correction coefficient.

[0061] S5: Repeat steps S1-S4 on multi-temporal infrared remote sensing images to generate angle-normalized time-series radiation datasets, thereby realizing the correction of changes in surface infrared radiation observed by remote sensing satellites.

[0062] Example 2: A system for correcting changes in infrared radiation on the Earth's surface observed by remote sensing satellites, comprising: Data preprocessing module: used to acquire multiple infrared band data and corresponding sensor observation zenith angles, and to perform radiometric calibration and atmospheric correction on the infrared band data to obtain surface radiance or surface temperature data. Model building module: Used to select pseudo-invariant feature points within the target area based on surface radiance or surface temperature data, and to establish a model of the relationship between sensor-observed zenith angle and surface radiation value based on the pseudo-invariant feature points; Corrected radiation calculation module: used to set the standard reference observation zenith angle and calculate the normalized corrected radiation value of the observed radiation value under different observation zenith angles to the standard reference observation zenith angle based on the relational model; Global Correction Module: This module applies normalized correction coefficients to the entire scene image, performing multiplicative correction on the original radiometric value of each pixel. Product generation module: This module is used to repeatedly perform the above processing on multi-temporal infrared remote sensing images to generate angle-normalized time-series radiation datasets, thereby realizing the correction of changes in infrared radiation on the Earth's surface observed by remote sensing satellites.

[0063] In one embodiment of the present invention, in view of the problem of measurement deviation of the surface infrared radiation caused by the deflection of the sensor and the change of the scanning angle during the observation process of the remote sensing satellite, an indoor multi-zenith angle simulation observation experiment is constructed to systematically observe the infrared radiation responses of different lithologies and multi-grain surface targets under the conditions of small-angle swing scanning and large-angle deflection. By precisely controlling the observation geometric relationship, the variation law of the radiation temperature at different observation angles is obtained, and the deviation between the actual zenith angle and the nominal observation angle and its directional characteristics are quantified, revealing the systematic radiation distortion mechanism caused by multi-angle observation. This experiment provides a direct physical basis and data support for the "correction scheme for surface infrared radiation change based on deflection angle normalization" proposed in the patent, verifying the necessity and feasibility of angle unification and normalization processing of remote sensing infrared data.

[0064] In actual infrared applications, there are mainly two observation methods: remote sensing satellites and drones. First, the remote sensing satellite observes the ground ( Figure 2 ), with a "m" - shaped swing observation. The observation angle variation range of the 4 observation lines (Line 1 - 4) is [-20°, 20°], which is called the small - angle observation experiment (abbreviated as E1 experiment). Second, the drone carries an infrared thermal imager for observation ( Figure 3 ), and the distribution of the measurement points is in the shape of ", and the observation angle variation range at 6 groups of observation points (Points 1 - 6) is , which is called the large - angle observation experiment (abbreviated as E2 experiment). Among them, the observation parameters of the small - angle and large - angle observation experiments are shown in Tables 1 and 2. In the tables is the height between the observation lens and the rock debris stage, is the horizontal distance between the observation lens and the rock debris stage, is the variation of the observation zenith angle.

[0065] Table 1 Statistical table of observation parameters for small - angle variation observation experiment

[0066] Table 2 Statistical table of observation parameters for large - angle variation observation experiment

[0067] In the small - angle variation observation experiment, the infrared thermal imager is fixed above the rock debris stage by a ladder. The rock debris stage is placed in a black box to isolate the influence of the surrounding background radiation. The black box is placed on a flatbed truck, and the flatbed truck is pushed along the calibrated position to form variation of the observation zenith angle; in the large - angle variation observation experiment, the observation height of the infrared thermal imager is adjusted by a lifting frame to form variation of the observation zenith angle.

[0068] Nine granularity rock debris samples are arranged in a matrix on the stage ( Figure 4 ), same observed zenith angle Actual zenith angle of different rock fragments Field of view deviation exists Actual zenith angle Calculation example as follows Figure 4 As shown, because the relative positions of the rock cuttings samples on the platform are fixed, at the same observation zenith angle... Rock cuttings of different sizes The deviation value is also determined (Table 3).

[0069] Table 3. Actual zenith angles corresponding to rock cuttings samples With the observed zenith angle Difference Statistics Table

[0070] Table 4 shows the grain size and number of different rock fragments. The infrared static observation experiment involved three types of lithology. Figure 5 This is a real-world image showing the arrangement of samples observed from multiple grain sizes of rocks. In the image, a represents marble; b represents granite; and c represents sandstone.

[0071] Table 4 Rock cuttings information and their corresponding numbers

[0072] Mineral composition, arrangement, and diagenesis significantly affect the radiation properties of rocks. Conducting orthogonal polarized light microscopy experiments to determine mineral composition and content is crucial for analyzing the infrared radiation properties of rocks.

[0073] Minerals and rocks ( Figure 6 It is a sedimentary rock, composed of medium to fine-grained minerals, and exhibits a massive structure. The grain size is 0.05–0.5 mm, with approximately 40% being 0.05–0.1 mm clastic particles and approximately 60% being 0.1–0.5 mm clastic particles. The main mineral components are quartz, feldspar, and other clastic fragments, as well as a small amount of muscovite.

[0074] The experiment was scheduled to take place after 10 PM. During the observation period, lights were turned off, movement was prohibited, and blackout curtains were drawn. The laboratory temperature was maintained at 22.1℃±0.1℃. The parameters of the thermal imager and lens are as follows: Infrared thermal imager: InfraTec 8325 mid-wave infrared thermal imager, with an observation band of 3.7 – 4.8 μm. The highest resolution of thermal images is 640 pixels × 512 pixels, with a maximum sampling rate of 120 P / s and a temperature recognition accuracy of 0.1K.

[0075] Standard lens: Model M83287, focal length 25 mm, focal length ratio 2.0, field of view (21×17)°. At an observation distance of 1 m, the maximum planar size of the observed object can reach 384 mm × 307 mm, the instantaneous field of view is 0.6 mrad, and the focusing range is from 0.3 m to infinity.

[0076] Infrared radiation temperature T Extracted from the defined area of ​​the thermal image ( Figure 7 Read rock cuttings #1 to #9. ( i Number the rock cuttings. i =1,2,......, 9; j Indicates the observation line. j =1, 2, 3, 4; k Indicates the observation angle. k =0°, 10°, -10°, 20°, -20°, 30°, 40°, 50°). The average infrared radiation temperature of rock fragments #1 to #9 was extracted from the thermal images. (Using a uniform box size and reducing boundary effects) is... value. Figure 7 R1 to R9 are the delineated areas of rock fragments #1 to #9, respectively.

[0077] The rock cuttings samples on the platform are arranged in a 3x3 grid, with fixed relative positions between them. The actual zenith angle faced by different rock cuttings samples within the same viewport is shown. The actual direction may differ slightly. It should be noted that the directional effect here refers to the observed zenith angle. In other words.

[0078] Using observations from a certain location Average of all observed azimuths difference Indicates, that is

[0079] In infrared static observation experiments, the directionality of strong indoor background radiation will cause the infrared radiation temperature in that direction to change. The direction that differs significantly from other directions is the direction with the dominant background radiation. Figures 8-10 They are marble, granite, and sandstone, respectively. Curves showing the change in observation angle, and the same observation zenith angle. A diagram illustrating the differences between different survey lines (i.e., azimuths).

[0080] Marble ( Figure 8 ): E1 ( Among them, Line 3 shows the most obvious difference from the other survey lines. All rock cuttings samples in Line 1, Rock cuttings #6 in Line 4 ( ), Rock cuttings #6 in Line 4 ( ), Rock cuttings #9 from Line 1 ( E2 and others also exhibit heterogeneity. Among the points, Points 2 shows the most significant difference from the other measurement points. Zhong 9# rock cuttings ( ), Middle No. 1 rock cuttings ( (These also exhibit diversity.)

[0081] granite( Figure 9 ): E1 ( Among them, Line 3 shows the most obvious difference from the other survey lines. Rock cuttings sample #1 from Line 4 ( ) and rock cuttings sample #2 ( ), Rock cuttings sample #1 from Line 4 ( ) and rock cuttings sample #2 ( E2 ( Among them, the heterogeneity of Points 6 is the most obvious. Rock cuttings sample #3 from Points 1 ( ), Points 3, No. 1 rock fragment sample ( The differences are also quite obvious.

[0082] sandstone( Figure 10 In E1 and E2, the rock fragments of various grain sizes exist It has good overall integrity and consistency. Rock cuttings sample #1 from Line 1 ( ), Rock cuttings #1 in Points 5 ( ), Rock cuttings sample #6 from Points 5 ( The differences are also quite obvious.

[0083] Based on the location markings of rock cuttings in Tables 1 and 2, the dominant direction of background radiation for the rock cuttings samples in this observation experiment was found to be directly below the rock cuttings stage.

[0084] The experiment involved effective shading and background radiation isolation, and different orientations. The difference was well controlled. Using the same observed zenith angle... Exploring the average values ​​from different directions. The angle effect. It should be noted that the angle effect here refers to the angle of observation. In terms of E1 experimental observations Change curve (left side) Figure 11 In the E2 experiment Curve graph (right side) Figure 11 ).

[0085] Marble ( Figure 11 a): Observed zenith angle (Left side image) The maximum value is 9#, and the minimum values ​​are 1# and 4#. The trend is "rising (1#→3#) → gradually decreasing (3#→4#) → stepping up (4#→9#)"; (Right side image) The maximum values ​​are 3#, 5#, and 9#, and the minimum values ​​are 1#, 4#, and 6#. The trend is "rising (1#→3#) → stepping up and down (3#→6#) → slight climbing (6#→9#)".

[0086] granite( Figure 11 (b) Observation of zenith angle (Left side image) The maximum value is 9#, and the minimum value is 1#. The overall trend is upward; (Right side image) Maximum value 1#, Minimum value 9# The trend is an overall step-like decline.

[0087] sandstone( Figure 11 c): Observed zenith angle (Left side image) Maximum values ​​3# and 9#, minimum value 1#. The trend is a slight increase in volatility. 。 (Right side image) The maximum values ​​are 3#, 5#, and 9#, and the minimum values ​​are 1# and 6#. The trend is "rapid increase (1#→3#) → slight fluctuation (3#→9#)".

[0088] This case study simulates multi-deflection angle observations by remote sensing satellites, systematically acquiring infrared radiation response data for targets of different lithologies and grain sizes under multi-zenith angle conditions. It quantitatively reveals the directional effects and angular distortion caused by changes in the observation deflection angle, verifying that surface infrared radiation exhibits a significant systematic shift characteristic with varying observation angles. Experimental results show that, without angle standardization, infrared data acquired under different observation conditions suffer from non-negligible radiation biases, which can easily be misinterpreted as actual surface changes.

[0089] In summary, the remote sensing infrared radiation variation correction method based on deflection angle normalization proposed in this invention can effectively eliminate systematic errors caused by observation geometric differences, improve the comparability and authenticity of multi-temporal and multi-angle remote sensing infrared data, and has clear experimental basis and engineering application value.

[0090] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of the invention.

Claims

1. A method for correcting changes in infrared radiation of the Earth's surface observed by remote sensing satellites, characterized in that, Includes the following steps: S1: Acquire multi-scene infrared band data and corresponding sensor observation zenith angles, and perform radiometric calibration and atmospheric correction on the infrared band data to obtain surface radiance or surface temperature data. S2: Based on surface radiance or surface temperature data, select pseudo-invariant feature points within the target area, and establish a model of the relationship between the sensor-observed zenith angle and the surface radiation value based on the pseudo-invariant feature points; S3: Set the standard reference observation zenith angle, and calculate the normalized radiance values ​​of the observed zenith angles to the standard reference observation zenith angle based on the relational model; S4: Apply the normalized correction coefficient to the entire scene image and perform multiplicative correction on the original radiometric value of each pixel; S5: Repeat steps S1-S4 on multi-temporal infrared remote sensing images to generate angle-normalized time-series radiation datasets, thereby realizing the correction of changes in surface infrared radiation observed by remote sensing satellites.

2. The method for correcting changes in infrared radiation of the Earth's surface observed by remote sensing satellites according to claim 1, characterized in that, The acquisition of multiple infrared band data and corresponding sensor observation zenith angles, followed by radiometric calibration and atmospheric correction of the infrared band data to obtain surface radiance or surface temperature data, includes: Multiple infrared band data from the same remote sensing satellite platform observed the same target area at different times are acquired. Radiometric calibration and atmospheric correction are performed on each data scene to obtain surface radiance or surface temperature data. At the same time, the sensor's zenith angle for each pixel in each data scene is extracted from the satellite auxiliary data.

3. The method for correcting changes in infrared radiation of the Earth's surface observed by remote sensing satellites according to claim 1, characterized in that, The model relating the zenith angle observed by the sensor to the Earth's surface radiation value is as follows: in, This represents the surface radiation value. For relational models, To enable sensors to observe the zenith angle from the ground, It represents the residual.

4. The method for correcting changes in infrared radiation of the Earth's surface observed by remote sensing satellites according to claim 3, characterized in that, The relation model The construction methods include: Cosine model based on the Lambertian approximation: in, This is the target radiation signal received by the sensor during vertical observation, and serves as the model's baseline reference value; A correction model based on atmospheric radiative transfer: a relational model is generated through simulation using simplified atmospheric radiative transfer equations. ; Statistical Nonparametric Fitting Model: Utilizing Pseudo-Invariant Feature Points The L-shaped scatter plot was used to obtain the relational model through nonparametric fitting using local weighted regression or spline functions. .

5. The method for correcting changes in infrared radiation of the Earth's surface observed by remote sensing satellites according to claim 1, characterized in that, The method for selecting pseudo-invariant feature points within the target area includes: using the stability of long-term vegetation index, surface reflectance, or nighttime light data for screening, and selecting several uniform pixels whose surface physical properties remain stable during the observation period as pseudo-invariant feature points.

6. The method for correcting changes in infrared radiation of the Earth's surface observed by remote sensing satellites according to claim 1, characterized in that, The formula for calculating the corrective radiation value is: in, To correct the radiation value, For arbitrary observation of zenith angle The observed radiation values ​​below, This is a model relating the standard reference observation zenith angle to the Earth's surface radiation value. For arbitrary observation of zenith angle A model relating to surface radiation values. For the standard reference observation zenith angle, For arbitrary observation of zenith angle The normalized correction coefficient.

7. The method for correcting changes in infrared radiation of the Earth's surface observed by remote sensing satellites according to claim 6, characterized in that, The normalized correction coefficient is applied to the entire image, and the original radiometric value of each pixel is multiplied for correction. The calculation formula is as follows: in, The corrected radiation value, The original radiation value. This is the normalized correction coefficient.

8. A system for correcting changes in infrared radiation observed from the Earth's surface by remote sensing satellites as described in any one of claims 1-7, characterized in that, include: Data preprocessing module: used to acquire multiple infrared band data and corresponding sensor observation zenith angles, and to perform radiometric calibration and atmospheric correction on the infrared band data to obtain surface radiance or surface temperature data. Model building module: Used to select pseudo-invariant feature points within the target area based on surface radiance or surface temperature data, and to establish a model of the relationship between sensor-observed zenith angle and surface radiation value based on the pseudo-invariant feature points; Corrected radiation calculation module: used to set the standard reference observation zenith angle and calculate the normalized corrected radiation value of the observed radiation value under different observation zenith angles to the standard reference observation zenith angle based on the relational model; Global Correction Module: This module applies normalized correction coefficients to the entire scene image, performing multiplicative correction on the original radiometric value of each pixel. Product generation module: This module is used to repeatedly perform the above processing on multi-temporal infrared remote sensing images to generate angle-normalized time-series radiation datasets, thereby realizing the correction of changes in infrared radiation on the Earth's surface observed by remote sensing satellites.

Images