Surface temperature inversion method based on thermal infrared remote sensing of unmanned aerial vehicle

Abstract

The invention relates to the technical field of surface temperature monitoring, and discloses an unmanned aerial vehicle thermal infrared remote sensing-based surface temperature inversion method, which comprises the following steps of S1, constructing a comprehensive flight parameter optimization model according to environmental conditions monitored in real time so as to calculate and set the optimal flight height and flight speed of an unmanned aerial vehicle; s2, in the flight process of the unmanned aerial vehicle, ground surface thermal infrared remote sensing data are collected through a thermal infrared sensor carried by the unmanned aerial vehicle, real-time flight parameters and environment parameters of the unmanned aerial vehicle are synchronously recorded, and by optimizing the flight parameters, the thermal infrared data quality stability is improved, and the inversion accuracy problem is solved; calibrating and optimizing model parameters by means of a ground target, and eliminating system errors without field calibration; and through multi-source correction, optimization inversion and noise reduction processing, high-resolution and high-precision surface temperature data is obtained, the multi-field high-precision requirement is met, and the thermal infrared remote sensing potential of the unmanned aerial vehicle is fully played.

Description

Technical Field

[0001] This invention relates to the field of land surface temperature monitoring technology, specifically a method for land surface temperature inversion based on UAV thermal infrared remote sensing. Background Technology

[0002] Land surface temperature is a crucial parameter in modern agriculture, environmental monitoring, and urban planning. Accurate acquisition of land surface temperature is essential for understanding surface energy exchange processes, assessing crop growth, and monitoring the urban heat island effect. Traditional methods for measuring land surface temperature primarily rely on ground-based meteorological stations and satellite remote sensing. However, ground-based meteorological stations are limited by their distribution, resulting in limited coverage. While satellite remote sensing offers broad coverage, its real-time performance and accuracy are affected by weather conditions and satellite transit times, requiring further improvement.

[0003] In recent years, the development of UAV technology has provided new means for monitoring land surface temperature. UAV thermal infrared remote sensing technology has shown great potential in acquiring land surface temperature due to its speed, flexibility, and high resolution. However, existing UAV thermal infrared remote sensing systems suffer from unstable image quality during data acquisition due to factors such as flight altitude, speed, and environmental conditions. This affects the accuracy of land surface temperature retrieval. Most systems use general or regional model parameters and lack real-time, on-site calibration methods specific to the flight mission and the specific area. This results in systematic errors in the retrieval results that are difficult to quantify, limiting the application of this technology in applications requiring high-precision quantitative analysis. Therefore, we propose a land surface temperature retrieval method based on UAV thermal infrared remote sensing to address the aforementioned problems. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method for land surface temperature inversion based on UAV thermal infrared remote sensing. This method solves the problems of unstable data acquisition quality caused by flight parameters and environmental influences in existing UAV thermal infrared remote sensing systems, as well as the lack of on-site model calibration for specific tasks and regions. Consequently, the land surface temperature inversion suffers from difficulties in quantifying systematic errors and cannot meet the requirements for high-precision quantitative analysis.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for retrieving land surface temperature based on UAV thermal infrared remote sensing, comprising the following steps:

[0006] S1. Based on the real-time monitored environmental conditions, construct a comprehensive flight parameter optimization model to calculate and set the optimal flight altitude and speed of the UAV;

[0007] S2. During the flight of the UAV, the UAV collects thermal infrared remote sensing data of the ground surface through its own thermal infrared sensor, and simultaneously records the UAV's real-time flight parameters and environmental parameters.

[0008] S3. Deploy several ground verification targets with known surface materials and emissivity within the UAV flight area, and simultaneously measure their actual surface temperature; ensure that the UAV flies over the targets and collects their thermal infrared data; based on the known properties of the targets and the synchronous measurement data, optimize the parameters of the atmospheric correction model and the surface emissivity estimation model.

[0009] S4. Using the atmospheric correction model parameters optimized by S3, atmospheric correction is performed on the collected thermal infrared remote sensing data. Combined with position and attitude correction based on flight parameters and sensor radiometric calibration correction, multi-source data fusion correction is completed.

[0010] S5. Using calibrated thermal infrared remote sensing data and combining it with the surface emissivity estimation model, surface temperature inversion calculation is performed.

[0011] S6. Spatial interpolation and noise reduction are performed on the surface temperature data obtained from the inversion calculation to generate a clear and accurate high-resolution surface temperature distribution map.

[0012] Preferably, in step S1, the integrated flight parameter optimization model is constructed in the following manner:

[0013] The air temperature, humidity and wind speed data of the flight area are obtained in real time using meteorological equipment, and combined with the terrain information of the area, and input into an optimization algorithm to calculate the optimal combination of flight altitude and speed.

[0014] The optimization algorithm aims to minimize the impact of flight attitude fluctuations on thermal infrared data acquisition and maximize the signal-to-noise ratio of the surface thermal radiation signal.

[0015] Preferably, in step S2, the flight parameters include the UAV's flight altitude, flight speed, pitch angle, roll angle, and yaw angle; the environmental parameters include atmospheric temperature, relative humidity, ground wind speed, and solar radiation intensity.

[0016] Preferably, in step S3, the ground verification targets are uniformly distributed within the area and include types with different specific emissivity, such as high, medium, and low.

[0017] The model parameter optimization involves using ground measurement data as the true values ​​and performing regression fitting on the parameters of the atmospheric correction and surface emissivity estimation models through the least squares method or Bayesian inversion algorithm.

[0018] Preferably, in step S4, the multi-source data fusion correction includes:

[0019] Atmospheric correction uses real-time collected atmospheric data to simulate and remove the absorption and emission effects of the atmosphere on thermal infrared signals through physical models.

[0020] Position and attitude correction: Based on the data recorded by the attitude measurement unit on the UAV, the thermal infrared image is rotated and its position adjusted to ensure that every point on the image corresponds precisely to the actual position on the ground.

[0021] Sensor radiation calibration is performed by linearly correcting the radiation response of the thermal infrared sensor by measuring a standard blackbody at a known temperature before and after flight.

[0022] Preferably, in step S5, the surface emissivity estimation model is as follows: based on multispectral thermal infrared data or visible-near infrared auxiliary data, a method combining vegetation cover and surface material classification is used to estimate the surface emissivity pixel by pixel; the surface temperature inversion adopts a split window algorithm or a single-channel algorithm, and introduces neighboring pixel temperature consistency constraints for optimization.

[0023] Preferably, in step S6, the spatial interpolation and noise reduction process includes: using the Kriging interpolation method to fill in the missing temperature regions, and combining wavelet transform to filter out noise from the temperature distribution map, so as to improve the spatial continuity and accuracy of the temperature data.

[0024] Beneficial effects

[0025] This invention provides a method for retrieving land surface temperature based on UAV thermal infrared remote sensing. Compared with existing technologies, it has the following advantages:

[0026] This land surface temperature retrieval method based on UAV thermal infrared remote sensing constructs a comprehensive flight parameter optimization model, combining real-time environmental and terrain information to determine the optimal flight altitude and speed. This effectively reduces the impact of flight attitude fluctuations on data acquisition, improves the stability of thermal infrared remote sensing image quality, and solves the problem of retrieval accuracy caused by unstable data quality. By deploying ground verification targets with different emissivity and simultaneously measuring temperatures, the atmospheric correction and land surface emissivity estimation model parameters are individually optimized, replacing general or regional model parameters and eliminating the difficult-to-quantify systematic errors caused by the lack of on-site calibration. After multi-source data fusion correction, optimized temperature retrieval algorithms, and spatial interpolation and noise reduction processing, high-resolution, high-precision land surface temperature data is finally obtained. This meets the needs of modern agriculture, environmental monitoring, urban planning, and other fields for high-precision quantitative analysis of land surface temperature, fully leveraging the potential of UAV thermal infrared remote sensing technology. Attached Figure Description

[0027] Figure 1 This is a flowchart of a method for retrieving land surface temperature based on UAV thermal infrared remote sensing, according to the present invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] like Figure 1 As shown:

[0030] A method for retrieving land surface temperature based on UAV thermal infrared remote sensing includes the following steps:

[0031] S1. Combine real-time environmental conditions and terrain information to construct a comprehensive flight parameter optimization model, calculate and set the optimal flight altitude and flight speed of the UAV;

[0032] The integrated flight parameter optimization model collects environmental data such as air temperature, humidity, and wind speed in the flight area in real time using meteorological equipment, and simultaneously imports terrain elevation data of the area to form the model input dataset. Parameter optimization is guided by two core objectives: first, minimizing the impact of flight attitude fluctuations on thermal infrared data acquisition to ensure the stability of sensor data acquisition; and second, maximizing the signal-to-noise ratio of the surface thermal radiation signal to improve the quality of the raw data. The input data is substituted into a preset optimization algorithm, and the optimal combination of flight altitude and speed is obtained through iterative calculation, providing parameter support for subsequent data acquisition.

[0033] S2. The UAV executes its flight mission according to the parameters set in S1, and simultaneously collects and records three types of data during the flight to ensure data integrity and spatiotemporal consistency.

[0034] The drone continuously collects surface thermal infrared remote sensing data using a thermal infrared sensor to capture surface thermal radiation information.

[0035] The drone's built-in attitude measurement unit records parameters such as flight altitude, flight speed, pitch angle, roll angle, and yaw angle in real time to reflect the flight status.

[0036] Simultaneously record environmental data such as atmospheric temperature, relative humidity, ground wind speed, and solar radiation intensity to provide a basis for subsequent model calibration;

[0037] S3. Deploy ground verification targets within the drone's flight coverage area, calibrate and optimize the core model parameters using on-site measurement data, and improve the model's adaptability. Deploy several ground verification targets using a uniform distribution principle. The targets must cover material types with high, medium, and low emissivity to ensure comprehensive coverage of the surface material characteristics of the target area.

[0038] While collecting thermal infrared data by drones flying over targets, high-precision temperature measurement equipment is used to measure the actual surface temperature of each target in the field to ensure the spatiotemporal synchronization of remote sensing data and measured data.

[0039] Using the true surface temperature and known emissivity of the target as the true values, the parameters of the atmospheric correction model and the surface emissivity estimation model are regressed and fitted using the least squares method or Bayesian inversion algorithm to obtain personalized model parameters that are suitable for the current flight mission and target area.

[0040] S4. Using the model parameters optimized in S3, combined with flight parameters and sensor calibration data, multi-dimensional fusion correction is performed on the collected thermal infrared remote sensing data to eliminate the influence of various error factors.

[0041] Atmospheric correction: Based on real-time collected environmental data such as atmospheric temperature and humidity, a physical model is used to simulate the absorption and emission process of thermal infrared signals by the atmosphere, thereby removing errors caused by atmospheric effects and restoring the true thermal radiation signal of the Earth's surface.

[0042] Position and attitude correction: Based on parameters such as pitch angle and roll angle recorded by the UAV attitude measurement unit, the thermal infrared image is subjected to geometric correction processing such as rotation and translation to ensure that each pixel on the image can accurately correspond to the actual position on the ground and eliminate geometric deviations caused by flight attitude fluctuations.

[0043] Sensor radiation calibration: Before and after the flight mission, the thermal infrared sensor is calibrated for radiation response using a standard blackbody of known temperature to establish a linear relationship between the sensor output signal and the actual radiation intensity, thereby eliminating systematic errors caused by the sensor's own response characteristics.

[0044] S5. Based on the corrected high-quality thermal infrared remote sensing data and combined with the optimized surface emissivity estimation model, a targeted algorithm is used to perform surface temperature inversion calculation.

[0045] Estimation of surface emissivity: Based on multispectral thermal infrared data or visible-near infrared auxiliary data, a method combining vegetation cover and surface material classification is used to estimate the surface emissivity pixel by pixel, fully taking into account the differences in radiation characteristics of different surface types.

[0046] Temperature inversion algorithm: Based on the data type and application scenario, either the split window algorithm or the single-channel algorithm is selected to calculate the land surface temperature. At the same time, the temperature consistency constraint of neighboring pixels is introduced to optimize the inversion process, avoid outliers in isolated pixels, and improve the rationality of the inversion results.

[0047] S6. Post-process the retrieved surface temperature data to fill in missing data areas and filter out noise, generating a clear and accurate high-resolution surface temperature distribution map.

[0048] By employing the Kriging interpolation method, the missing temperature areas caused by observation blind spots and sensor failures are filled in using surrounding effective temperature data, ensuring the spatial continuity of temperature data. Combined with wavelet transform technology, random noise in the temperature data is filtered out. While preserving the core features of temperature distribution, the smoothness and accuracy of the data are improved, and finally, a high-resolution, high-reliability surface temperature distribution map is output.

[0049] In this implementation plan: a multi-rotor UAV equipped with a high-resolution thermal infrared sensor is selected, along with a high-precision attitude measurement unit, a portable weather station, a high-precision thermometer, and a standard blackbody. Simultaneously, ground verification targets with different emissivity are fabricated, including high emissivity targets (such as black rubber sheets), medium emissivity targets (such as gray cement sheets), and low emissivity targets (such as aluminum sheets). Each target measures 1m x 1m and has a smooth, non-reflective surface. Next, topographic elevation data of the target area is collected to delineate the flight area boundaries and plan flight routes.

[0050] One hour before the start of the flight mission, air temperature, humidity and wind speed data of the target area are collected using a portable weather station. Combined with terrain elevation data, the data are input into a comprehensive flight parameter optimization model. The optimal flight altitude and speed are obtained through iterative calculation using a genetic algorithm.

[0051] The drone was launched according to the optimized parameters. The thermal infrared sensor collected thermal infrared data of the ground surface, the attitude measurement unit recorded parameters such as flight altitude, speed, and pitch angle, and the meteorological equipment recorded environmental parameters in real time. All data was stored to the ground control terminal through the wireless transmission module to ensure that the data timestamps were consistent.

[0052] Five to ten ground verification targets are evenly deployed in the flight area, with a distance of no less than 50m between adjacent targets. When the UAV flies over the targets, ground personnel use a high-precision thermometer to measure and record the actual surface temperature of the targets. After the flight, the parameters of the atmospheric correction model and the surface emissivity estimation model are fitted and optimized using the least squares method with the actual measured temperature of the targets and the known emissivity as the true values.

[0053] Using optimized atmospheric correction parameters and real-time environmental data, atmospheric correction is performed using the MODTRAN atmospheric radiative transfer model. Based on attitude measurement data, an affine transformation algorithm is used to correct the position and attitude of the thermal infrared image. Based on standard blackbody calibration data, a sensor radiation response equation is established to complete the radiation calibration correction, and finally, the corrected thermal infrared radiation data is obtained.

[0054] The normalized vegetation index (NDVI) is calculated based on multispectral thermal infrared data to obtain vegetation cover. Combined with the surface material classification results, the surface emissivity is estimated pixel by pixel. The split window algorithm is used to calculate the preliminary surface temperature data by substituting the corrected thermal radiation data and the estimated emissivity and introducing the temperature consistency constraint of neighboring pixels.

[0055] Ordinary Kriging interpolation was used to fill in the missing areas in the preliminary temperature data, with the interpolation search radius set to 3 pixels. Wavelet transform was used to decompose the temperature data into three levels, and the data was reconstructed after removing high-frequency noise components, finally generating a surface temperature distribution map with a spatial resolution ≥0.5m.

[0056] This solution constructs a comprehensive flight parameter optimization model, combining real-time environmental and terrain information to determine the optimal flight altitude and speed. This effectively reduces the impact of flight attitude fluctuations on data acquisition, improves the stability of thermal infrared remote sensing image quality, and solves the problem of inversion accuracy caused by unstable data quality. By deploying ground verification targets with different emissivity and simultaneously measuring temperatures, the parameters of the atmospheric correction and surface emissivity estimation models are individually optimized, replacing general or regional model parameters and eliminating the difficult-to-quantify systematic errors caused by the lack of on-site calibration. After multi-source data fusion correction, optimized temperature inversion algorithms, and spatial interpolation and noise reduction processing, high-resolution, high-precision surface temperature data is finally obtained. This meets the needs of modern agriculture, environmental monitoring, urban planning, and other fields for high-precision quantitative analysis of surface temperature, fully leveraging the potential of UAV thermal infrared remote sensing technology.

[0057] It should be noted that: ground verification targets should be set up in an area without obstruction or shadow. When measuring the temperature of the target, the vertical distance between the thermometer probe and the target surface should not exceed 0.5m. Each target should be measured 3 times and the average value should be taken to ensure the accuracy of the measured data.

[0058] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for retrieving land surface temperature based on UAV thermal infrared remote sensing, characterized in that: Includes the following steps: S1. Based on the real-time monitored environmental conditions, construct a comprehensive flight parameter optimization model to calculate and set the optimal flight altitude and speed of the UAV; S2. During the flight of the UAV, the UAV collects thermal infrared remote sensing data of the ground surface through its own thermal infrared sensor, and simultaneously records the UAV's real-time flight parameters and environmental parameters. S3. Deploy several ground verification targets with known surface materials and emissivity within the UAV flight area, and simultaneously measure their actual surface temperature; ensure that the UAV flies over the targets and collects their thermal infrared data; based on the known properties of the targets and the synchronous measurement data, optimize the parameters of the atmospheric correction model and the surface emissivity estimation model. S4. Using the atmospheric correction model parameters optimized by S3, atmospheric correction is performed on the collected thermal infrared remote sensing data. Combined with position and attitude correction based on flight parameters and sensor radiometric calibration correction, multi-source data fusion correction is completed. S5. Using calibrated thermal infrared remote sensing data and combining it with the surface emissivity estimation model, surface temperature inversion calculation is performed. S6. Spatial interpolation and noise reduction are performed on the surface temperature data obtained from the inversion calculation to generate a clear and accurate high-resolution surface temperature distribution map.

2. The method for retrieving land surface temperature based on UAV thermal infrared remote sensing according to claim 1, characterized in that: In S1, the integrated flight parameter optimization model is constructed in the following way: The air temperature, humidity and wind speed data of the flight area are obtained in real time using meteorological equipment, and combined with the terrain information of the area, and input into an optimization algorithm to calculate the optimal combination of flight altitude and speed. The optimization algorithm aims to minimize the impact of flight attitude fluctuations on thermal infrared data acquisition and maximize the signal-to-noise ratio of the surface thermal radiation signal.

3. The method for retrieving land surface temperature based on UAV thermal infrared remote sensing according to claim 1, characterized in that: In S2, the flight parameters include the UAV's flight altitude, flight speed, pitch angle, roll angle, and yaw angle; the environmental parameters include atmospheric temperature, relative humidity, ground wind speed, and solar radiation intensity.

4. The method for retrieving land surface temperature based on UAV thermal infrared remote sensing according to claim 1, characterized in that: In S3, the ground verification targets are uniformly distributed within the area and include types with different specific emissivity, such as high, medium, and low. The model parameter optimization involves using ground measurement data as the true values ​​and performing regression fitting on the parameters of the atmospheric correction and surface emissivity estimation models through the least squares method or Bayesian inversion algorithm.

5. The method for retrieving land surface temperature based on UAV thermal infrared remote sensing according to claim 1, characterized in that: In step S4, the multi-source data fusion correction includes: Atmospheric correction uses real-time collected atmospheric data to simulate and remove the absorption and emission effects of the atmosphere on thermal infrared signals through physical models. Position and attitude correction: Based on the data recorded by the attitude measurement unit on the UAV, the thermal infrared image is rotated and its position adjusted to ensure that every point on the image corresponds precisely to the actual position on the ground. Sensor radiation calibration is performed by linearly correcting the radiation response of the thermal infrared sensor by measuring a standard blackbody at a known temperature before and after flight.

6. The method for retrieving land surface temperature based on UAV thermal infrared remote sensing according to claim 1, characterized in that: In S5, the surface emissivity estimation model is as follows: based on multispectral thermal infrared data or visible-near infrared auxiliary data, a method combining vegetation cover and surface material classification is used to estimate the surface emissivity pixel by pixel; the surface temperature inversion adopts a split window algorithm or a single-channel algorithm, and introduces neighboring pixel temperature consistency constraints for optimization.

7. The method for retrieving land surface temperature based on UAV thermal infrared remote sensing according to claim 1, characterized in that: In step S6, the spatial interpolation and noise reduction process includes: using the Kriging interpolation method to fill in the missing temperature regions, and combining wavelet transform to filter out noise from the temperature distribution map, so as to improve the spatial continuity and accuracy of the temperature data.

Images