Sea surface temperature inversion method based on unmanned aerial vehicle non-real-time radiation calibration infrared data
By preprocessing and mapping modeling UAV infrared data, the problem of lack of real-time radiometric calibration for high-sensitivity UAV infrared imagers was solved, enabling high-precision inversion of sea surface temperature and improving data quality and usability.
Patent Information
- Application Number
- CN202511242118.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-02
- Publication Date
- 2025-12-12
AI Technical Summary
The lack of a real-time radiometric calibration system in the high-sensitivity infrared imager of UAVs makes it impossible to quantitatively invert sea surface temperature using traditional methods, resulting in insufficient data quality and usability.
By preprocessing infrared data to eliminate noise and non-uniformity, performing georegistration and standardization, establishing a mapping relationship model between integration time and relative temperature difference, and constructing a sea surface temperature inversion model using synchronous reference points, the sea surface temperature at each point within the observation area is calculated.
This achievement enables the inversion of sea surface temperature from high-sensitivity infrared data from UAVs, improving data quality and availability, and solving the problem of lacking real-time radiometric calibration. It has significant theoretical and practical implications.
Smart Images
Figure CN121113271A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of marine environmental remote sensing information processing technology, and more specifically, to a method for sea surface temperature inversion based on unmanned aerial vehicle (UAV) non-real-time radiometric calibration infrared data. Background Technology
[0002] Sea surface temperature, as the primary interface parameter for ocean-atmosphere interaction, plays an irreplaceable role in climate systems, marine ecology, and human activities. In particular, as an indicator of ocean dynamic processes such as ocean fronts, mesoscale eddies, and stratification stability, it has significant economic and military applications in marine fisheries, shipping, typhoon intensity prediction, sonar detection, and amphibious warfare.
[0003] Currently, the global sea surface temperature observation network includes Argo, satellites, and ships, generating a large amount of data daily that provides effective support for marine scientific research, disaster early warning, and military applications. Spaceborne infrared radiometers offer advantages such as wide-area, long-term, and fixed-route observations, enabling the acquisition of sea surface temperature information across the entire ocean area and at all times. However, they are affected by lighting conditions and cloud cover at sea, resulting in limited effective imaging windows and low spatiotemporal resolution and poor timeliness in fused products. Argo and ship-based observation data offer high precision, but the data is sparse and costly. Long-endurance UAV high-sensitivity infrared imagers offer advantages such as mobility, rapid response, low cost, high resolution, and long revisit cycles, playing a crucial role in real-time marine environmental monitoring and refined marine management in key sea areas.
[0004] Sea surface temperature retrieval methods based on infrared radiometers include single-channel algorithms, multi-channel nonlinear algorithms, and split-window algorithms. These algorithms are all based on real-time calibration of infrared radiometers, ensuring operational reliability. For a specific long-endurance UAV with a high-sensitivity infrared imager employing a single 8-10 micrometer band and pixel-level digital infrared imaging technology, ultra-high system sensitivity is achieved through a pixel-level digital integration mechanism. This provides the relative temperature difference corresponding to a specific integration time. However, without a real-time radiometric calibration system on board, absolute radiometric calibration is impossible. Therefore, traditional physical models and statistical methods cannot be used to quantitatively retrieve sea surface temperature. There is an urgent need to establish a model relating integration time and relative temperature difference to solve the problem of retrieving sea surface temperature based on high-sensitivity infrared data from UAVs. Summary of the Invention
[0005] To address the aforementioned problems, the present invention aims to provide a sea surface temperature inversion method based on UAV-based non-real-time radiometric calibration infrared data, thereby solving the challenge of inverting sea surface temperature using high-sensitivity UAV infrared data.
[0006] To achieve the above technical objectives, this application provides a sea surface temperature inversion method based on unmanned aerial vehicle (UAV) non-real-time radiometric calibration infrared data, comprising the following steps: To improve data quality and usability, the UAV infrared data is preprocessed; Based on the preprocessed infrared data, the relative temperature difference at the actual selected integration time is calculated. According to the selected synchronization reference point, the sea surface temperature of the synchronization reference point in the observation area is obtained, and a sea surface temperature inversion model is established to calculate the sea surface temperature at each point in the observation area.
[0007] Preferably, when preprocessing UAV infrared data, the preprocessing process is completed by eliminating periodic stripe noise and isolated point noise, as well as non-uniformity caused during imaging, and by performing georegistration and standardization.
[0008] Preferably, during georeferencing, the geographic coordinates of the top left, top right, bottom right, and bottom left corner points, data rows and columns, and corresponding resolutions recorded in the original data file are used to calculate the geographic coordinates, and the original data is projected onto the geographic grid.
[0009] Preferably, during standardization, the nearest neighbor interpolation method combined with a small window filter is used to perform unified resolution processing on multi-resolution images.
[0010] Preferably, when calculating the relative temperature difference at the actual selected integration time, the relative temperature difference of the observed records at the actual selected integration time is calculated by establishing a mapping relationship model between the integration time and the relative temperature difference of the observed records, based on the known relative temperature difference corresponding to the specific integration time and the observed records.
[0011] Preferably, when obtaining the sea surface temperature of the synchronous reference point within the observation area, the sea surface temperature of the synchronous reference point is expressed as:
[0012] In the formula, To superimpose the cold skin effect and the solar warming effect on sea surface temperature, The sea surface temperature at 1m, measured by AXCTD or buoy. Temperature difference caused by the cold skin effect The temperature difference is caused by the solar warming effect.
[0013] Preferably, when constructing the sea surface temperature inversion model, the sea surface temperature inversion model is expressed as:
[0014] in, For the observation data area Sea surface temperature at the point, Reference point for the observation area sea surface temperature, They are respectively , Observational records of the point Observation records when the integration time is selected The corresponding relative temperature difference.
[0015] This invention discloses a sea surface temperature inversion system based on UAV-based non-real-time radiometric calibration infrared data, used to implement the aforementioned sea surface temperature inversion method based on UAV-based non-real-time radiometric calibration infrared data. The system includes: The data processing module is used to preprocess the UAV infrared data to improve data quality and usability; The sea surface temperature inversion module is used to obtain the sea surface temperature of the synchronous reference point in the observation area based on the preprocessed infrared data by calculating the relative temperature difference at the actual selected integration time and according to the selected synchronous reference point. It also establishes a sea surface temperature inversion model to calculate the sea surface temperature at each point in the observation area.
[0016] The present invention discloses the following technical effects: This invention effectively solves the problems of lack of real-time radiometric calibration and susceptibility to noise and radiation distortion in high-sensitivity infrared payload data from UAVs. A mapping model between the integration time and the relative temperature difference of the observed values is established based on a specific integration time. Using the sea surface temperature of a synchronous reference point, sea surface temperature inversion is achieved from high-sensitivity infrared data from UAVs without real-time radiometric calibration, which has significant theoretical and practical value. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is the inversion flowchart described in this invention; Figure 2 This is a diagram illustrating the noise removal effect described in this invention; Figure 3 This is an image showing the effect of isolated noise removal as described in this invention; Figure 4 This is a georeferencing effect diagram as described in this invention; Figure 5 This is a schematic diagram of the reference point acquisition based on UAV synchronization as described in this invention; Figure 6This is a schematic diagram of obtaining a synchronization reference point based on a shipborne infrared radiometer as described in this invention; Figure 7 This is a diagram illustrating the sea surface temperature inversion effect described in this invention. Detailed Implementation
[0019] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.
[0020] like Figures 1-7 As shown, this invention provides a sea surface temperature inversion method based on high-sensitivity, non-real-time radiometric calibration infrared data from unmanned aerial vehicles (UAVs), comprising the following steps: S1. High-sensitivity infrared data preprocessing for UAVs eliminates periodic stripe noise and isolated point noise caused by their own factors, as well as non-uniformity caused during the imaging process, and performs georegistration and standardization to improve data quality and usability.
[0021] Step S1 specifically includes preprocessing such as removing periodic stripe noise and isolated point noise from the high-sensitivity infrared data of the UAV, addressing non-uniformity caused during the imaging process, georegistration, and standardization.
[0022] (1) Noise Removal Remove periodic stripe noise and isolated point noise caused by the infrared load itself.
[0023] 1) Periodic stripe noise removal First, locate the stripe noise in the image; second, take the abnormal row and replace the abnormal row value with the average of the three rows of data before and after it (non-abnormal rows); finally, return the image after noise removal.
[0024] 2) Isolated point noise filtering Isolated noise points are located by calculating the image gradient value to pinpoint isolated points with significant gradient changes. The average gray value of the neighboring windows of the isolated point is then used to fill the isolated noise point, and the image after removing the noise from the isolated points is returned.
[0025] (2) Non-uniformity correction The non-uniformity problem caused during the imaging process is mainly manifested as higher brightness in the center of the image and lower brightness at the edges. This can be corrected by using the relationship between the observation values of the reference radiation source and the detector unit.
[0026] ; in, For a certain illuminance Next, the The output value of each detector unit These are the values after non-uniformity correction. , It is a correction coefficient calculated using the response output of the reference radiation source and each detector unit.
[0027] (3) Geographic registration Using the geographic coordinates of the top left, top right, bottom right, and bottom left corner points, data rows and columns, and corresponding resolutions recorded in the original data file, the geographic coordinates are calculated using bilinear interpolation, and the original data is projected onto the geographic grid.
[0028] Given the four corner points: top left Top right Bottom right Bottom left Coordinate calculation for any point The coordinate interpolation steps are as follows: First, using the coordinates of four points... Linear interpolation is performed in the direction, i.e.: ; ; Then, in Linear interpolation in the direction: ; (4) Standardization process Due to varying flight requirements and the influence of sea conditions and weather, the UAVs fly at different altitudes, even within the same flight, resulting in differences in the swath width and resolution of the infrared payload data. To fully utilize flight data from different altitudes, image resolution is standardized. A nearest neighbor interpolation method combined with a small window filter is used to unify the resolution of multi-resolution images. The formula for the nearest neighbor interpolation method is as follows: ; ; in, and The coordinates in the reference image, and These are the coordinates after zooming in or out. and It is the width of the reference image and the target image. and It represents the height of the reference image and the target image.
[0029] S2. Based on the relative temperature difference corresponding to a specific integration time and observed records, establish a mapping relationship model between integration time and relative temperature difference, and calculate the relative temperature difference at the actual selected integration time.
[0030] Step S2 specifically includes establishing a mapping relationship model between the integration time and the relative temperature difference of the observed recorded values based on the known specific integration time and the relative temperature difference corresponding to the observed recorded values, and calculating the relative temperature difference of the observed recorded values at the actual selected integration time.
[0031] The specific time integral of the infrared load calibration is known to be (unit is) The relative temperature difference between the observed value DN and the observed value DN (Unit is) ), in the interval The fitting formula for the relative temperature difference corresponding to the observed value DN when the integration time is selected is: ; In the formula: Select the actual integration time The relative temperature difference corresponding to the time record value DN is expressed in units of 1000 kJ / m². .
[0032] S3. Select a synchronization reference point and obtain the sea surface temperature of the synchronization reference point within the observation area.
[0033] Step S3 specifically includes selecting a synchronization reference point and obtaining the sea surface temperature of the synchronization reference point within the observation area.
[0034] (1) Acquisition of synchronous reference points The acquisition of the synchronous reference point can be achieved by simultaneously measuring the temperature and depth of the drone with a temperature and depth measurement device (AXCTD or buoy), or by using a shipborne infrared radiometer to conduct synchronous observations within the drone's flight area.
[0035] (2) Acquisition of sea surface temperature at synchronous reference points When using a shipborne infrared radiometer to obtain a synchronization reference point, since both measure the sea surface temperature at a depth of 10–20 μm, the sea surface temperature measured by the shipborne infrared radiometer can be directly used as the sea surface temperature of the synchronization reference point.
[0036] When using contact temperature sensors such as AXCTD and buoys for on-site measurements, since these sensors measure sea surface temperature at a certain depth (approximately 1m), to improve the accuracy of the sea surface temperature at the reference point, the effects of the cold skin effect and solar warming effect can be superimposed on the sea surface temperature measured by the AXCTD or buoy. This transforms the sea surface temperature into the same depth as the infrared payload observation, expressed as: ; In the formula, To superimpose the cold skin effect and the solar warming effect on sea surface temperature, The sea surface temperature at 1m, measured by AXCTD or buoy. Temperature difference caused by the cold skin effect The temperature difference is caused by the solar warming effect.
[0037] S4. Establish a sea surface temperature inversion model based on synchronous reference points and calculate the sea surface temperature at each point in the observation area.
[0038] Step S4 specifically includes establishing a sea surface temperature inversion model based on a synchronous reference point and calculating the sea surface temperature at each point within the observation area.
[0039] The high-sensitivity infrared imager of the UAV without real-time calibration system observes the sea surface at a selected integration time. The sea surface temperature inversion formula is as follows: ; in, For the observation data area Sea surface temperature at the point, Reference point for the observation area sea surface temperature, They are respectively , Observational records of the point Observation records when the integration time is selected The corresponding relative temperature difference.
[0040] S5. Sea surface temperature inversion accuracy assessment: Compare with synchronous satellite remote sensing and marine observation data, and use statistical standard deviation to test and assess the inversion accuracy.
[0041] Step S5 specifically includes comparing the sea surface temperature inversion results of the UAV high-sensitivity non-real-time calibration system infrared data with the sea surface temperature of matching points other than the reference point in the synchronous satellite remote sensing data, reanalysis data, and marine measurement data of contact measurement equipment (buoys, CTD, XBT, AXCTD, etc.), and evaluating the inversion accuracy by statistical standard deviation.
[0042] Example: A sea surface temperature retrieval method based on UAV high-sensitivity non-real-time radiometric calibration infrared imager is implemented using data images. The specific process is as follows: Figure 1 As shown.
[0043] (1) Infrared data preprocessing: noise removal, radiometric correction, georegistration and standardization are performed on an image of observation data from a high-sensitivity infrared imager of a UAV.
[0044] 1) Noise Removal ① Periodic stripe noise removal The first step is to locate the stripe noise in the image; The second step is to take the abnormal row and replace the abnormal row value with the average of the three rows of data before and after it (non-abnormal rows). The third step involves visualizing the image after noise removal, as shown in the image below. Figure 2 As shown.
[0045] ② Isolated point noise removal The first step is to locate isolated noise points by calculating the image gradient values and locating isolated points with large gradient value changes. The second step, filling isolated noise points, can be done by taking the average gray value of the nearby windows. The third step involves visualizing the image after isolating and removing noise, as shown in the image below. Figure 3 As shown.
[0046] 2) Non-uniformity correction The first step is to select a reference radiation source whose irradiance is known; The second step involves using the reference radiation source as the observation target and performing multiple measurements. Then, using the least squares method, the non-uniformity correction coefficients are calculated by fitting the response output of the reference radiation source and each detector unit. , ; The third step is to perform radiometric correction on the detection data. ; in, For a certain illuminance Next, the The output value of each detector unit These are the values after non-uniformity correction. , This is the non-uniformity correction coefficient.
[0047] 3) Geographic registration Using the geographic coordinates of the top left, top right, bottom right, and bottom left corners, the data rows and columns, and the corresponding resolution recorded in the original data, the original data is projected onto the geographic grid. The schematic diagram of the original data and the geographic registration grid is shown in Figure 4.
[0048] Given the four corner points: top left Top right Bottom right Bottom left Coordinate calculation for any point The coordinate interpolation steps are as follows: First, using the coordinates of four points... Linear interpolation is performed in the direction, i.e.: ; ; Then, in Linear interpolation in the direction: ; 4) Standardization Processing Nearest neighbor interpolation is used to unify the resolution of multi-resolution images. The calculation formula is: ; ; in, and The coordinates in the reference image, and These are the coordinates after zooming in or out. and It is the width of the reference image and the target image. and It represents the height of the reference image and the target image.
[0049] (2) Relative temperature difference mapping model The specific time integral of the infrared load calibration is known to be (unit is) The relative temperature difference between the observed value DN and the observed value DN (Unit is) When ), then in the interval The fitting formula for the relative temperature difference corresponding to the observed value DN when the integration time is selected is: ; In the formula: Select the actual integration time The relative temperature difference corresponding to the time record value DN is expressed in units of 1000 kJ / m². .
[0050] (3) Acquisition of synchronous reference points 1) Selection of Synchronization Reference Point The acquisition of a synchronization reference point can be achieved by simultaneously measuring temperature and depth using a temperature and depth measurement device (AXCTD or buoy) mounted on a drone, such as... Figure 5 As shown, shipborne infrared radiometers can also be used for synchronous observation within the UAV's flight area, such as... Figure 6 As shown.
[0051] 2) Acquisition of sea surface temperature at synchronous reference points The observed data first undergoes data preprocessing to remove outliers. Sea surface temperatures at depths of 10–20 μm measured by both the shipborne infrared radiometer and the UAV infrared imager can be directly used as the sea surface temperature at the synchronization reference point. In the field, contact temperature sensors such as AXCTDs and buoys measure sea surface temperatures at a certain depth (approximately 1 m). To improve the accuracy of the reference point's sea surface temperature, the effects of cold skin effect and solar warming effect can be superimposed on the sea surface temperature measured by the AXCTD or buoy, converting it to the sea surface temperature at the same depth as the infrared imager observation, expressed as: ; In the formula, To superimpose the cold skin effect and the solar warming effect on sea surface temperature, The sea surface temperature at 1m, measured by AXCTD or buoy. Temperature difference caused by the cold skin effect The temperature difference is caused by the solar warming effect.
[0052] (4) Sea surface temperature inversion The high-sensitivity infrared imager of the UAV without real-time calibration system observes the sea surface at a selected integration time. The sea surface temperature inversion formula is as follows: ; in, For the observation data area Sea surface temperature at the point, Reference point for the observation area sea surface temperature, They are respectively , Observational records of the point The relative temperature difference at the selected integration time. Figure 7 The result is the sea surface temperature inversion at the top left corner reference point with a temperature of 32.13°C at 24 minutes of integration time.
[0053] This invention provides a sea surface temperature retrieval method for high-sensitivity infrared imager data from long-endurance unmanned aerial vehicles (UAVs) without real-time radiometric calibration systems. The method involves preprocessing the high-sensitivity infrared data, including noise removal, non-uniformity correction, and georegistration. A mapping model between integration time and relative temperature difference is constructed based on a specific integration time and corresponding relative temperature difference to obtain the relative temperature difference at the actual selected integration time. A sea surface temperature retrieval model based on a synchronous reference point is established to achieve sea surface temperature retrieval from high-sensitivity infrared data from UAVs without real-time radiometric calibration systems. Synchronous sea surface temperature data is collected to verify the accuracy of the sea surface temperature retrieval.
[0054] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0055] In the description of this invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.
[0056] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. A method for retrieving sea surface temperature from non-real-time radiometrically calibrated infrared data based on an unmanned aerial vehicle, characterized in that, The method comprises the following steps: In order to improve the data quality and availability, the unmanned aerial vehicle infrared data is preprocessed; Based on the preprocessed infrared data, the relative temperature difference at the actual selected integration time is calculated, the sea surface temperature of the synchronous reference point in the observation area is obtained according to the selected synchronous reference point, and a sea surface temperature inversion model is established for calculating the sea surface temperature of each point in the observation area.
2. The sea surface temperature inversion method based on unmanned aerial vehicle non-real-time radiation calibration infrared data according to claim 1, wherein: When preprocessing the unmanned aerial vehicle infrared data, periodic stripe noise and isolated point noise are eliminated, non-uniformity caused in the imaging process is eliminated, and geographic registration and standardization processing are performed to complete the preprocessing process.
3. The sea surface temperature inversion method based on unmanned aerial vehicle non-real-time radiation calibration infrared data according to claim 2, wherein: When performing geographic registration, the geographic coordinates of the upper left, upper right, lower right and lower left corners, the data rows and columns and the corresponding resolutions recorded in the original data file are used to calculate the geographic coordinates by using the bilinear interpolation method, and the original data is projected onto the geographic grid.
4. The sea surface temperature inversion method based on unmanned aerial vehicle non-real-time radiation calibration infrared data according to claim 3, wherein: When performing standardization processing, the nearest neighbor interpolation method is used in combination with a small window filter to perform uniform resolution processing on the multi-resolution image.
5. The sea surface temperature inversion method based on unmanned aerial vehicle non-real-time radiation calibration infrared data according to claim 4, wherein: When calculating the relative temperature difference at the actual selected integration time, the relative temperature difference of the observation record value at the actual selected integration time is calculated by establishing a mapping relationship model between the integration time and the relative temperature difference of the observation record value according to the known specific integration time and the corresponding relative temperature difference of the observation record value.
6. The sea surface temperature inversion method based on unmanned aerial vehicle non-real-time radiation calibration infrared data according to claim 5, wherein: When obtaining the sea surface temperature of the synchronous reference point in the observation area, the sea surface temperature of the synchronous reference point is represented as:
7. where, SST is the sea surface temperature, SST is the sea surface temperature measured by AXCTD or buoy at 1 m, SST is the sea surface temperature measured by AXCTD or buoy at 1 m, SST is the sea surface temperature measured by AXCTD or buoy at 1 m, 8. The sea surface temperature inversion method based on unmanned aerial vehicle non-real-time radiation calibration infrared data according to claim 6, wherein: When constructing the sea surface temperature inversion model, the sea surface temperature inversion model is represented as:
9. wherein, to observe the data area sea surface temperature of the point, to observe the reference point of the area sea surface temperature of the point, respectively , observation record value of the point, observation record when the integral time is selected corresponding relative temperature difference.
10. A sea surface temperature retrieval system based on unmanned aerial vehicle non-real-time radiometrically calibrated infrared data, for implementing the sea surface temperature retrieval method based on unmanned aerial vehicle non-real-time radiometrically calibrated infrared data according to claim 1, characterized in that, It comprises: a data processing module for preprocessing the unmanned aerial vehicle infrared data to improve the data quality and availability; a sea surface temperature inversion module for calculating the relative temperature difference at the actual selected integration time based on the preprocessed infrared data, obtaining the sea surface temperature of the synchronous reference point in the observation area according to the selected synchronous reference point, and establishing a sea surface temperature inversion model for calculating the sea surface temperature of each point in the observation area.