Integrated satellite remote sensing and video detection of comprehensive fire monitoring and early warning method and platform

Abstract

This invention relates to a comprehensive fire monitoring and early warning method and platform integrating satellite remote sensing and video detection, comprising: generating a two-dimensional thermal radiation map based on satellite remote sensing data; generating a three-dimensional spatial scene based on multi-angle video monitoring data; mapping the two-dimensional thermal radiation map to the corresponding surface location in the three-dimensional spatial scene, and performing spatiotemporal extrapolation based on the time difference between video and satellite data and the fire spread pattern to generate a three-dimensional thermal radiation field; performing bidirectional atmospheric attenuation correction on the three-dimensional thermal radiation field and video light signal using real-time meteorological data; spatially overlaying and trend analysis of the corrected thermal radiation anomaly area and the video fire and smoke characteristic area, and determining whether it is a real fire based on the analysis and evaluation results; if it is determined to be a real fire, generating an early warning. This invention, through the fusion of satellite remote sensing and video detection, can accurately determine the actual fire situation, significantly improve the accuracy of fire monitoring and early warning, and ensure fire prevention and early warning.

Description

Technical Field

[0001] This invention relates to the field of fire monitoring and early warning technology, and more specifically, to a comprehensive fire monitoring and early warning method and platform that integrates satellite remote sensing and video detection. Background Technology

[0002] The core requirement of fire monitoring and early warning technology lies in achieving a balance between wide-area coverage and real-time accurate identification. Currently, satellite remote sensing and video detection are two mainstream technologies. Satellite remote sensing, with its wide-area observation advantage, can quickly cover large areas and capture information on abnormal surface thermal radiation, providing basic data support for large-scale fire investigation. Video detection, on the other hand, has strong real-time characteristics, dynamically capturing the visual characteristics of fire and smoke in local areas, aiding in close-range fire monitoring. However, both technologies have inherent limitations when applied individually: satellite remote sensing has a limited observation cycle, insufficient temporal resolution, and a fixed observation time difference, making it difficult to synchronously reflect real-time dynamic changes in the fire situation; video detection, affected by factors such as deployment range and terrain obstruction, has limited monitoring coverage and cannot meet the fire monitoring needs of large areas such as forests and grasslands.

[0003] As the requirements for monitoring accuracy and response time in fire prevention and control continue to increase, the limitations of existing technologies are becoming increasingly apparent. Due to the differences in data characteristics and observation dimensions between satellite remote sensing and video detection, the two technologies lack an effective collaborative mechanism, preventing them from complementing each other's advantages. This results in practical fire monitoring either failing to capture the real-time development of a fire due to reliance on satellite remote sensing data, or failing to achieve large-scale, comprehensive monitoring due to reliance on video detection. Furthermore, problems such as spatiotemporal asynchrony and environmental interference faced by different data sources cannot be effectively resolved, leading to insufficient accuracy in fire identification, a high risk of misjudgment and missed detection, and difficulty in forming timely and reliable fire warnings, posing a significant challenge to early fire prevention and response. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a comprehensive fire monitoring and early warning method and platform that integrates satellite remote sensing and video detection, addressing the problems existing in the prior art.

[0005] The technical solution adopted by this invention to solve its technical problem is: to construct a comprehensive fire monitoring and early warning method integrating satellite remote sensing and video detection, comprising:

[0006] Radiometric calibration is performed on satellite remote sensing data of the target area to generate a two-dimensional thermal radiation map with geographic coordinates;

[0007] Acquire multi-angle video surveillance data of the target monitoring area, and perform three-dimensional scene reconstruction based on the multi-angle video surveillance data to generate a three-dimensional spatial scene including terrain and equipment location;

[0008] The two-dimensional thermal radiation map is mapped to the corresponding surface location of the three-dimensional spatial scene, and spatiotemporal extrapolation is performed based on the time difference between the video and the satellite and the fire spread pattern to generate a three-dimensional thermal radiation field synchronized with the video time.

[0009] The three-dimensional thermal radiation field and video light signal are subjected to bidirectional atmospheric attenuation correction using real-time meteorological data.

[0010] The corrected thermal radiation anomaly area and the video fire and smoke feature area are spatially overlaid and their changing trends are analyzed. Based on the analysis and evaluation results, it is determined whether it is a real fire. If it is determined to be a real fire, an early warning is generated.

[0011] In the integrated fire monitoring and early warning method combining satellite remote sensing and video detection described in this invention, the step of performing radiometric calibration on the satellite remote sensing data of the target area to generate a two-dimensional thermal radiation map with geographic coordinates includes:

[0012] Based on the atmospheric-surface radiation coupling relationship, atmospheric correction is performed on the satellite remote sensing data to obtain the apparent radiation value at the top of the atmosphere;

[0013] Identify the cloud interference region in the apparent radiation value of the top of the atmosphere, estimate the theoretical radiation of the ground surface under the cloud based on the surface thermal inertia and satellite transit time, replace the cloud interference region, and obtain the cloud area corrected thermal radiation value.

[0014] The corrected thermal radiation value of the cloud area is combined with the real-time acquired aerosol optical thickness data to perform aerosol scattering effect compensation and remove pseudo-high temperature points caused by specular reflection caused by solar flare angle to obtain the two-dimensional thermal radiation map.

[0015] In the integrated fire monitoring and early warning method combining satellite remote sensing and video detection described in this invention, the atmospheric-surface radiation coupling relationship is obtained through the following method:

[0016] Based on the analysis of historical clean day data observed over a long period in the target area, the localized synergistic variation relationship between surface emissivity and surface temperature was obtained.

[0017] The real-time meteorological data acquired synchronously during satellite transit is analyzed to extract the water vapor content in the atmosphere of the target area.

[0018] The localized coordinated change relationship and the water vapor content are substituted into the physical process describing thermal radiation transfer, and combined with satellite observation geometric parameters, the corresponding atmospheric path radiation and atmospheric transmittance are simulated and calculated to determine the atmospheric-surface radiation coupling relationship.

[0019] In the integrated fire monitoring and early warning method combining satellite remote sensing and video detection described in this invention, the step of performing three-dimensional scene reconstruction based on the multi-angle video surveillance data to generate a three-dimensional spatial scene including terrain and equipment locations includes:

[0020] From the multi-angle video surveillance data, extract a set of visual feature points containing terrain edges and the outlines of fixed facilities within the target area;

[0021] Based on synchronously acquired lidar point cloud data, spatial geometric correction is performed on the visual feature point set to obtain a sparse three-dimensional point cloud.

[0022] Based on the constraint of terrain continuity, spatial interpolation is performed on the sparse 3D point cloud to obtain a dense 3D point cloud;

[0023] The dense 3D point cloud is triangularly meshed to construct a surface, and vegetation texture information extracted from the multi-angle video surveillance data is fused to generate the 3D spatial scene.

[0024] In the integrated fire monitoring and early warning method combining satellite remote sensing and video detection described in this invention, the step of mapping the two-dimensional thermal radiation map to the corresponding surface location of the three-dimensional spatial scene, and performing spatiotemporal extrapolation based on the time difference between the video and the satellite and the fire spread pattern to generate a three-dimensional thermal radiation field synchronized with the video time includes:

[0025] Based on the geometry of the surface triangular facets of the three-dimensional spatial scene and the geometry of satellite imaging, a projection relationship from the pixels of the thermal radiation map to the three-dimensional surface facets is established.

[0026] The two-dimensional thermal radiation map is corrected pixel by pixel using the projection relationship to determine the target triangular facet corresponding to the radiation value of each pixel.

[0027] Based on the geometric deformation ratio between the target triangular facet in three-dimensional space and the actual pixel projection, the radiation value is normalized by area to obtain a set of thermal radiation values ​​mapped to the surface location of the three-dimensional space scene.

[0028] Based on the set of thermal radiation values ​​mapped to the surface location of the three-dimensional spatial scene, the abnormal high temperature zone is determined;

[0029] Based on the synchronously acquired wind speed and direction data and time difference, the boundary expansion and radiation intensity changes of the abnormal high temperature zone from satellite time to video time are deduced, and the deduced fire state at the video time is obtained.

[0030] The simulated fire scene state at the video moment is assigned to the corresponding position in the three-dimensional spatial scene to form the three-dimensional thermal radiation field synchronized with the video time.

[0031] In the integrated fire monitoring and early warning method combining satellite remote sensing and video detection described in this invention, the step of performing bidirectional atmospheric attenuation correction on the three-dimensional thermal radiation field and video light signal using real-time meteorological data includes:

[0032] Key atmospheric parameters are determined based on the real-time meteorological data; these key atmospheric parameters include atmospheric aerosol optical thickness and atmospheric water vapor content.

[0033] Based on the atmospheric radiative transfer relationship and the aforementioned key atmospheric parameters, atmospheric transmittance and path radiative intensity are calculated respectively.

[0034] The three-dimensional thermal radiation field is attenuated and compensated according to the atmospheric transmittance to obtain the corrected thermal radiation field.

[0035] Background removal is performed on the video optical signal based on the path radiation intensity to obtain the corrected optical signal.

[0036] In the integrated fire monitoring and early warning method combining satellite remote sensing and video detection described in this invention, the step of determining key atmospheric parameters based on the real-time meteorological data includes:

[0037] Acquire real-time observation data uploaded by a ground-based monitoring network deployed in the target area; the observation data includes: aerosol concentration and atmospheric water vapor concentration;

[0038] Simultaneously extract the apparent reflectance information of the satellite remote sensing data in the visible and near-infrared bands;

[0039] The observation data and the apparent reflectance information are fused together, and the atmospheric aerosol optical thickness and the atmospheric water vapor content are obtained by coupling the different response characteristics of the two to atmospheric conditions.

[0040] In the integrated fire monitoring and early warning method combining satellite remote sensing and video detection described in this invention, the step of spatially overlaying and analyzing the changing trends of the corrected thermal radiation anomaly area and the video fire and smoke feature area includes:

[0041] Extract the thermal radiation anomaly region from the corrected thermal radiation field;

[0042] Extract the video smoke and fire feature region from the corrected optical signal;

[0043] Calculate the overlap area between the thermal radiation anomaly area and the video fire and smoke feature area in the three-dimensional spatial scene;

[0044] Construct a time-varying sequence of radiation intensity in the thermal radiation anomaly zone and a time-varying sequence of brightness in the video fire and smoke feature zone.

[0045] In the integrated fire monitoring and early warning method combining satellite remote sensing and video detection described in this invention, the step of determining whether a fire is real based on analysis and evaluation results, and generating an early warning if a real fire is determined, includes:

[0046] Analyze the spatial stability of the overlapping region;

[0047] Compare the trends of the time-varying radiant intensity sequence and the time-varying luminance sequence;

[0048] Based on the trend assessment results and the aforementioned spatial stability, the overall confidence level for the occurrence of the fire was determined;

[0049] Based on the comprehensive confidence level, it is determined whether it is a real fire. If it is determined to be a real fire, an early warning is generated.

[0050] This invention also provides a comprehensive fire monitoring and early warning platform integrating satellite remote sensing and video detection, comprising:

[0051] The satellite data preprocessing module is used to perform radiometric calibration on satellite remote sensing data of the target area and generate a two-dimensional thermal radiation map with geographic coordinates.

[0052] The video 3D reconstruction module is used to acquire multi-angle video surveillance data of the target monitoring area, and perform 3D scene reconstruction based on the multi-angle video surveillance data to generate a 3D spatial scene including terrain and equipment location;

[0053] The spatiotemporal fusion simulation module is used to map the two-dimensional thermal radiation map to the corresponding surface location of the three-dimensional spatial scene, and to perform spatiotemporal simulation based on the time difference between the video and the satellite and the fire spread pattern to generate a three-dimensional thermal radiation field synchronized with the video time.

[0054] A multi-source data correction module is used to perform bidirectional atmospheric attenuation correction on the three-dimensional thermal radiation field and video light signal using real-time meteorological data.

[0055] The intelligent analysis and early warning module is used to spatially overlay and analyze the changing trends of the corrected thermal radiation anomaly area and the video fire and smoke feature area, and to determine whether it is a real fire based on the analysis and evaluation results. If it is determined to be a real fire, an early warning is generated.

[0056] The integrated fire monitoring and early warning method and platform combining satellite remote sensing and video detection of the present invention has the following beneficial effects:

[0057] This invention accurately maps a two-dimensional thermal radiation map generated from satellite remote sensing data onto a three-dimensional spatial scene, and combines the fire spread pattern with the time difference between satellite and video to complete spatiotemporal extrapolation, achieving time synchronization between the thermal radiation field and video monitoring. At the same time, through spatial overlay and change trend analysis of the corrected thermal radiation anomaly area and the video fire and smoke feature area, the invention accurately locates the real fire situation, significantly improving the spatiotemporal matching degree of fire monitoring and the accuracy of early warning judgment, ensuring that fires can be identified in a timely manner and trigger effective early warnings.

[0058] Meanwhile, the precise correction of atmospheric interference and cloud cover effects during satellite data preprocessing, combined with bidirectional attenuation compensation based on real-time meteorological parameters, effectively eliminates multiple environmental interferences, ensuring the authenticity and purity of core monitoring data. Furthermore, the high-precision 3D scene generated by video 3D reconstruction, in conjunction with pixel-by-pixel terrain correction and area normalization processing in spatiotemporal fusion simulation, further improves the accuracy of matching thermal radiation data with surface morphology, making the monitoring of fire location and extent more precise and providing reliable and detailed data support for fire response. Attached Figure Description

[0059] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings:

[0060] Figure 1 This is a flowchart illustrating the integrated fire monitoring and early warning method combining satellite remote sensing and video detection provided in this embodiment of the invention.

[0061] Figure 2 This is a flowchart illustrating the three-dimensional thermal radiation field generation method provided in an embodiment of the present invention;

[0062] Figure 3 This is a schematic flowchart of the bidirectional atmospheric attenuation correction method provided in an embodiment of the present invention;

[0063] Figure 4 This is a schematic flowchart of the real fire analysis and judgment method provided in the embodiments of the present invention;

[0064] Figure 5 This is a logical block diagram of the integrated fire monitoring and early warning platform that integrates satellite remote sensing and video detection provided in the embodiments of the present invention. Detailed Implementation

[0065] 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.

[0066] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “said” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0067] Depending on the context, the word "if" or "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0068] Furthermore, the timing of the steps in the following method embodiments is merely an example and not a strict limitation.

[0069] refer to Figure 1 , Figure 1 This paper illustrates a preferred embodiment of the integrated fire monitoring and early warning method combining satellite remote sensing and video detection provided by the present invention.

[0070] Specifically, such as Figure 1 As shown, the integrated fire monitoring and early warning method combining satellite remote sensing and video detection includes steps S10, S20, S30, S40, and S50. The specific execution of each step is as follows:

[0071] Step S10: Perform radiometric calibration on the satellite remote sensing data of the target area to generate a two-dimensional thermal radiation map with geographic coordinates.

[0072] In some embodiments, performing radiometric calibration on satellite remote sensing data of the target area to generate a two-dimensional thermal radiation map with geographic coordinates includes: performing atmospheric correction on the satellite remote sensing data based on the atmospheric-surface radiation coupling relationship to obtain the apparent radiation value at the top of the atmosphere; identifying cloud interference areas in the apparent radiation value at the top of the atmosphere, estimating the theoretical radiation of the surface below the clouds based on surface thermal inertia and satellite transit time, replacing the cloud interference areas, and obtaining the cloud-corrected thermal radiation value; and performing aerosol scattering effect compensation on the cloud-corrected thermal radiation value, combined with real-time acquired aerosol optical thickness data, and removing specular reflection pseudo-high temperature points caused by solar flare angles to obtain the two-dimensional thermal radiation map. This two-dimensional thermal radiation map is a geographically coordinated two-dimensional thermal radiation map used for mapping.

[0073] In this embodiment of the invention, the atmospheric-surface radiation coupling relationship is obtained through the following methods: Analysis of historical clean day data from long-term observations of the target area yields a localized synergistic variation relationship between surface emissivity and surface temperature; real-time meteorological data synchronously acquired during satellite transit is analyzed to extract the water vapor content in the atmosphere of the target area; the localized synergistic variation relationship and water vapor content are substituted into the physical process describing thermal radiation transfer, and combined with satellite observation geometric parameters, the corresponding atmospheric path radiation and atmospheric transmittance are simulated and calculated to determine the atmospheric-surface radiation coupling relationship.

[0074] Preferably, the simulation calculation of atmospheric path radiation and atmospheric transmittance can be achieved using a simplified atmospheric radiative transfer model, which can be specifically expressed as:

[0075] ;

[0076] In the formula, Indicates atmospheric transmittance in the thermal infrared band. Indicates the first The absorption coefficients of atmospheric components, such as water vapor, ozone, and carbon dioxide, in the thermal infrared band depend on atmospheric pressure. and temperature The cross-section, Indicates the first The column content of various atmospheric absorption components in the vertical direction. This represents the zenith angle observed by the satellite. This represents the cosecant value of the zenith angle, used to convert the content of vertical columns to the content on an inclined path.

[0077] In practice, historical clean-up day data consists of selected satellite observations with stable land cover types and no cloud cover or aerosol interference. Continuous observation periods cover the land surface conditions of the target area at different times of different seasons. By statistically analyzing the variation of land surface emissivity with land surface temperature under the same land cover type, a one-to-one correlation curve is established between the two. It should be noted that the aforementioned localized synergistic change relationship is a qualitative / regular description of the inherent relationship between the two. In this embodiment, the one-to-one correlation curve is a concrete manifestation of the localized synergistic change relationship quantified and visualized through statistical methods.

[0078] Specifically, real-time meteorological data comes from observations at ground-based meteorological stations precisely matched to satellite transit times. By extracting key elements such as relative humidity, temperature, and air pressure from the meteorological data, the atmospheric column water vapor content of the target area is obtained using the conversion rules between humidity and water vapor content. Here, the atmospheric column water vapor content is the specific parameter name for the water vapor content in the atmosphere.

[0079] Specifically, the satellite observation geometric parameters are calculated from satellite ephemeris data and the geographic coordinates of the target area. The physical process of thermal radiation transfer is the process by which thermal radiation emitted from the Earth's surface is absorbed and scattered by the atmosphere before reaching the satellite sensor. By combining the established correlation curves and the obtained water vapor content, the correspondence between the degree of atmospheric absorption and scattering of thermal radiation and surface radiation is clarified. This correspondence is the atmospheric-surface radiation coupling relationship.

[0080] Specifically, based on the specific correlation values ​​of atmospheric transmittance and atmospheric path radiation in the atmospheric-surface radiation coupling relationship, the original radiation value received by the satellite sensor is deducted from the influence of atmospheric path radiation, and then divided by the atmospheric transmittance to obtain the corrected apparent radiation value of the top of the atmosphere.

[0081] In this embodiment of the invention, the identification of cloud interference areas specifically involves: by comparing the spatial distribution characteristics of apparent radiance values ​​at the top of the atmosphere, areas with radiance values ​​significantly lower than those of the surrounding land surface are designated as cloud interference areas. Based on the inherent physical properties determined by the land cover type of the target area, and combined with the time information of the satellite's transit time, the surface thermal inertia is used to calculate the surface thermal state at that moment, thereby estimating the theoretical radiance value of the land surface beneath the clouds. The estimated theoretical radiance value is then used to replace the apparent radiance value of the cloud interference area, resulting in the cloud-corrected thermal radiation value.

[0082] In this embodiment of the invention, the aerosol optical thickness data comes from real-time observation data from ground-based aerosol observation stations. The thermal radiation value is compensated for by subtracting scattered radiation based on the correlation between aerosol optical thickness and scattered radiation. The solar flare angle is the angle between the satellite observation direction and the solar incident direction. By judging whether the observation angle corresponding to the thermal radiation value is within the range that is prone to specular reflection, pseudo-high temperature points that meet the range are removed from the thermal radiation data.

[0083] Specifically, the atmospheric transmittance in the thermal infrared band is obtained by calculating the negative exponent of the cumulative value of the product of the absorption coefficients of atmospheric absorbing components and the content of vertical columns. The absorption coefficient of a certain atmospheric absorbing component in the thermal infrared band depends on the profiles of atmospheric pressure and temperature. Its value is determined by simulating different atmospheric pressure and temperature conditions in a laboratory environment, measuring the degree of absorption of the corresponding absorbing component to the thermal infrared radiation, and establishing the correspondence between the absorption coefficient and atmospheric pressure and temperature.

[0084] Specifically, no. The column content of atmospheric absorption components in the vertical direction is obtained from real-time observation data of atmospheric composition observation satellites. The zenith angle observed by the satellite is calculated from satellite ephemeris data and geographic coordinates of the target area. The cosecant value of the zenith angle is obtained by trigonometric function conversion.

[0085] Specifically, the atmospheric radiative transfer model clarifies the atmosphere's ability to transmit radiation in the thermal infrared band. The trend is that the larger the absorption coefficient of the atmospheric absorbing components, the higher the vertical column content, and the larger the satellite observation zenith angle, the smaller the atmospheric transmittance in the thermal infrared band, and vice versa.

[0086] It should be noted that those skilled in the art can also directly map the raw radiation data acquired by satellite sensors into a two-dimensional thermal radiation map by combining it with geographic coordinate information, without performing any preprocessing operations such as atmospheric correction, cloud interference removal, or aerosol compensation; or rely on the single-point thermal radiation data collected by ground-deployed thermal infrared monitoring equipment to generate a large-scale two-dimensional thermal radiation map through spatial interpolation algorithms such as Kriging interpolation and inverse distance weighting.

[0087] Overall, this solution presents a preferred implementation method that establishes a localized relationship between surface emissivity and temperature based on historical clean day data of the target area, and accurately derives the atmospheric-surface radiation coupling relationship by combining real-time meteorological data, thereby eliminating atmospheric influences at their source.

[0088] Furthermore, in response to cloud interference, the theoretical radiation of the ground surface under the clouds is scientifically estimated by analyzing the correlation between surface thermal inertia and satellite transit time, thus avoiding the information loss caused by directly discarding cloud area data in traditional methods.

[0089] More importantly, by simultaneously carrying out aerosol scattering compensation and eliminating false high-temperature points at the solar flare angle, the problem of multiple interference superposition is completely solved. The resulting thermal radiation map has three major characteristics: high precision, high continuity, and strong anti-interference, which can provide reliable data support for fire monitoring and early warning.

[0090] Step S20: Obtain multi-angle video surveillance data of the target monitoring area, and perform three-dimensional scene reconstruction based on the multi-angle video surveillance data to generate a three-dimensional spatial scene including terrain and equipment location.

[0091] In some embodiments, performing 3D scene reconstruction based on multi-angle video surveillance data to generate a 3D spatial scene including terrain and equipment locations includes: extracting a set of visual feature points containing terrain edges and the outlines of fixed facilities within the target area from the multi-angle video surveillance data; performing spatial geometric correction on the visual feature point set based on synchronously acquired LiDAR point cloud data to obtain a sparse 3D point cloud; performing spatial interpolation on the sparse 3D point cloud based on terrain continuity constraints to obtain a dense 3D point cloud; constructing a triangular mesh surface on the dense 3D point cloud and fusing vegetation texture information extracted from the multi-angle video surveillance data to generate a 3D spatial scene. Here, the sparse 3D point cloud is a sparse 3D point cloud with precise geodetic coordinates; the dense 3D point cloud is a dense 3D point cloud representing the continuous morphology of the land surface; and the 3D spatial scene is a 3D spatial scene including terrain undulations and surface attributes of objects on the land surface.

[0092] In practice, visual feature point sets containing terrain edges and the outlines of fixed facilities within the target area are extracted from multi-angle video surveillance data. The video images from each angle are traversed frame by frame to capture areas where grayscale values ​​change abruptly. These areas correspond to terrain edges and the outlines of fixed facilities. Points with stable positions and high recognizability in each angle video are then selected and integrated to form a unified set of visual feature points.

[0093] Furthermore, based on the synchronously acquired lidar point cloud data, spatial geometric correction is performed on the visual feature point set. The lidar point cloud data and video surveillance data are aligned with the same timestamp. Corresponding points in the visual feature point set and lidar point cloud are found. Using the precise geodetic coordinates of the lidar point cloud as a reference, the spatial position of the visual feature points is adjusted to eliminate the geometric deviation caused by visual imaging, and a sparse three-dimensional point cloud with precise geodetic coordinates is obtained.

[0094] Furthermore, based on the constraint of terrain continuity, spatial interpolation is performed on the sparse three-dimensional point cloud. Following the principle that the elevation changes of adjacent terrain points conform to the natural undulation law, the reasonable elevation values ​​of adjacent points are calculated based on the known points of the sparse three-dimensional point cloud to fill the gaps between the point clouds and obtain a dense three-dimensional point cloud that represents the continuous morphology of the land surface.

[0095] Finally, a triangular mesh surface is constructed from the dense 3D point cloud. Points are connected according to their spatial adjacency to form a triangular mesh, thus building the basic surface structure of the land. At the same time, vegetation texture information is extracted from the video and the texture is accurately fitted to the corresponding triangular mesh area to generate a 3D spatial scene that includes terrain undulations and surface attributes of objects on the land.

[0096] It should be noted that those skilled in the art can also use multi-view stereo images acquired by satellites to extract the contour features of terrain and features, calculate the terrain elevation through stereo matching technology, and construct a three-dimensional spatial scene by combining the geographic annotation information of the equipment; or deploy ground-based lidar equipment in the target area to directly scan and acquire point cloud data of terrain and equipment, and construct a three-dimensional spatial scene after denoising processing, without the need to combine video data.

[0097] Overall, this solution, as a preferred implementation, solves the problem of blind spots in pure LiDAR scanning by extracting refined visual feature point sets of terrain and equipment from multi-angle video, ensuring the integrity of scene details. It corrects visual feature points based on the precise geodetic coordinates of the synchronous LiDAR point cloud, eliminating geometric distortion in video imaging and ensuring the coordinate accuracy of the 3D scene. Furthermore, spatial interpolation based on terrain continuity constraints fills the gaps in the sparse point cloud, achieving continuous representation of terrain morphology, while the fusion of vegetation textures enhances the realism of the scene. This solution does not require large-scale equipment deployment, balancing the needs of wide-area coverage and refined modeling. It can efficiently generate real-time, high-precision 3D spatial scenes, providing accurate spatial data support for fire monitoring and early warning.

[0098] Step S30: Map the two-dimensional thermal radiation map to the corresponding surface location in the three-dimensional spatial scene, and perform spatiotemporal simulation based on the time difference between the video and the satellite and the fire spread pattern to generate a three-dimensional thermal radiation field synchronized with the video time.

[0099] In a preferred embodiment, such as Figure 2 As shown, step S30 specifically includes the following sub-steps:

[0100] Sub-step S301: Based on the geometry of the surface triangular patch and satellite imaging geometry of the three-dimensional spatial scene, establish the projection relationship from the pixels of the thermal radiation map to the three-dimensional surface patch;

[0101] Sub-step S302: Perform pixel-by-pixel terrain correction on the two-dimensional thermal radiation map using projection relationships to determine the target triangular facet corresponding to the radiation value of each pixel;

[0102] Sub-step S303: Based on the geometric deformation ratio between the target triangular facet in three-dimensional space and the actual pixel projection, the radiation value is normalized by area to obtain a set of thermal radiation values ​​mapped to the surface location of the three-dimensional space scene.

[0103] Sub-step S304: Determine the abnormal high temperature zone based on the set of thermal radiation values ​​mapped to the surface location of the three-dimensional space scene;

[0104] Sub-step S305: Based on the synchronously acquired wind speed and direction data and time difference, extrapolate the boundary expansion and radiation intensity changes of the abnormal high temperature zone from the satellite time to the video time, and obtain the extrapolated fire state at the video time.

[0105] Sub-step S306: Assign the simulated fire scene state at the moment of the video to the corresponding position in the three-dimensional spatial scene to form a three-dimensional thermal radiation field synchronized with the video time.

[0106] In practice, the vertex coordinates and orientation information of all triangular facets on the surface in the three-dimensional spatial scene are extracted. Then, the orbital parameters, observation angle and imaging range of the satellite imaging are obtained. By associating the spatial position of the triangular facets with the pixel coordinate system of the satellite imaging, a one-to-one correspondence rule between each thermal radiation map pixel and the three-dimensional surface facet is constructed to form a stable projection relationship.

[0107] Furthermore, each pixel of the two-dimensional thermal radiation map is traversed according to the constructed projection relationship, and the corresponding three-dimensional ground surface patch is matched according to the coordinate position of the pixel. At the same time, the pixel projection deviation caused by terrain undulation is corrected to ensure that the radiation value of each pixel accurately corresponds to the target triangular facet in the three-dimensional spatial scene.

[0108] Finally, the ratio of the actual spatial area of ​​the target triangular facet to the projected area of ​​the corresponding pixel is calculated as the geometric deformation ratio. Then, the radiation value of each pixel is multiplied by this ratio to eliminate the radiation value distortion caused by the projection deformation. The corrected radiation values ​​are then organized according to the position of the triangular facet to form a set of thermal radiation values.

[0109] In this embodiment of the invention, the deduction of the boundary expansion of the abnormal high-temperature zone and the change in radiation intensity can be based on the following physical model of fire spread:

[0110] ;

[0111] In the formula, Indicates the moment in the video. The spatial location vectors derived from the deduction, representing key feature points at the fire boundary, Indicates the time of satellite transit The identified initial spatial position vector, The base spread rate is an empirical or semi-empirical parameter that is related to the type of surface combustibles and humidity. The wind acceleration factor is a dimensionless empirical coefficient that characterizes the effect of wind on the spread of fire. Represents the wind speed vector. This represents the angle between the wind direction vector and the direction vector of the fire spread normal. This represents the integral variable, and time is represented.

[0112] In practice, the radiation values ​​of each triangular facet in the thermal radiation value set are compared with the normal surface thermal radiation benchmark value of the target area. Areas with radiation values ​​exceeding the benchmark value and conforming to the high temperature characteristics of the early stage of a fire are designated as abnormal high temperature zones, and the initial boundary and radiation intensity distribution of the abnormal high temperature zones are clarified.

[0113] Furthermore, the basic spread rate was obtained by recording the fire boundary expansion distance per unit time through fire spread experiments simulating different types of surface combustibles and humidity conditions in the target area. The wind acceleration factor was determined by calculating the wind's contribution coefficient to the spread rate through multiple sets of fire spread control experiments with different wind speeds and directions. The wind speed vector was obtained from synchronously observed ground meteorological data, and the angle between the wind direction vector and the fire line spread normal direction vector was calculated by the geometric relationship between the boundary direction of the abnormal high temperature zone in the three-dimensional spatial scene and the wind direction.

[0114] Then, by combining the time difference between the satellite and the video, the positional changes of key feature points at the fire boundary and the attenuation or enhancement of radiation intensity are gradually calculated in each time interval to form the simulated fire state at the moment of the video.

[0115] Through the above simulation method, the actual development state of the fire field within the time difference between satellite and video is accurately restored, providing accurate data support for the subsequent construction of three-dimensional thermal radiation field. Among them, the larger the basic spread rate, the faster the fire field expands; the larger the wind acceleration factor, the stronger the wind speed's role in promoting spread; the closer the angle between the wind direction and the fire line spread normal direction is to zero degrees, the more obvious the wind's driving effect on spread; and the radiation intensity changes synchronously with the fire field expansion range and is affected by the distribution of combustibles.

[0116] Finally, the state information such as the fire boundary position and radiation intensity distribution obtained from the video moment is accurately matched to the corresponding triangular facets of the three-dimensional spatial scene, so that each triangular facet carries the corresponding thermal radiation attribute, and finally integrated to form a three-dimensional thermal radiation field that is completely synchronized with the video time and has both spatial form and thermal radiation characteristics.

[0117] It should be noted that those skilled in the art can also use other methods, such as obtaining multi-period two-dimensional thermal radiation maps of satellite transit, calculating the thermal radiation data within the time difference between the satellite and the video through linear interpolation, and then directly superimposing it onto the three-dimensional spatial scene to generate the corresponding thermal radiation field; or relying on the thermal imaging function of video surveillance, inverting the thermal radiation value within the monitoring range, and then matching the data to a local area of ​​the three-dimensional spatial scene.

[0118] Overall, this scheme, as a preferred implementation method, achieves accurate mapping from two-dimensional thermal radiation maps to the three-dimensional Earth surface by utilizing the projection relationship between satellite imaging geometry and triangular facets of the three-dimensional scene. Combined with area normalization processing, it eliminates projection distortion errors, ensuring spatial matching accuracy. Simultaneously, based on real-time data such as wind speed and direction and the physical laws of fire spread, it conducts spatiotemporal simulations to accurately reconstruct the boundary expansion and radiation intensity changes of the fire field within the time difference between satellite and video data. The generated three-dimensional thermal radiation field possesses both spatial accuracy and temporal synchronization.

[0119] This solution not only solves the spatial adaptation problem of satellite data, but also makes up for the lack of dynamic simulation in other methods, providing reliable dynamic spatial data support for fire monitoring and early warning.

[0120] Step S40: Perform bidirectional atmospheric attenuation correction on the three-dimensional thermal radiation field and video light signal using real-time meteorological data.

[0121] In a preferred embodiment, such as Figure 3 As shown, step S40 specifically includes the following sub-steps:

[0122] Step S401: Determine key atmospheric parameters based on real-time meteorological data; key atmospheric parameters include: atmospheric aerosol optical thickness and atmospheric water vapor content;

[0123] Step S402: Based on the atmospheric radiative transfer relationship and key atmospheric parameters, calculate the atmospheric transmittance and path radiative intensity respectively;

[0124] The atmospheric transmittance is the atmospheric transmittance for the thermal infrared band; the path radiance is the path radiance for the visible / near-infrared band.

[0125] Among them, based on atmospheric radiative transfer relationships and key atmospheric parameters, atmospheric transmittance and path radiative intensity are calculated respectively. The core involves the physical decomposition of the signals received by the satellite sensor, which can be achieved through the following modes:

[0126] ;

[0127] In the formula, Indicates the satellite or video sensor in the band Total radiance received, Indicates the Earth's surface in the band The surface emissivity or surface reflectivity, This represents the temperature of a blackbody at the Earth's surface calculated using Planck's law. band radiance, Indicates along the observed zenith angle Atmospheric transmittance of the path, Indicates along the observed zenith angle Atmospheric path radiation of the path.

[0128] Step S403: Perform attenuation compensation on the three-dimensional thermal radiation field based on atmospheric transmittance to obtain the corrected thermal radiation field;

[0129] Step S404: Remove the background from the video light signal based on the path radiation intensity to obtain the corrected light signal.

[0130] In this embodiment, determining key atmospheric parameters based on real-time meteorological data includes: acquiring observation data uploaded in real-time by a ground-based monitoring network deployed in the target area; the observation data includes aerosol concentration and atmospheric water vapor concentration; simultaneously extracting apparent reflectance information from satellite remote sensing data in the visible and near-infrared bands; fusing the ground-based observation data with the apparent reflectance information, and inverting the atmospheric aerosol optical thickness and atmospheric water vapor content by coupling the different response characteristics of the two to atmospheric conditions.

[0131] In practice, the observation data uploaded in real time by the ground-based monitoring network deployed in the target area is obtained. The observation data includes aerosol concentration and atmospheric water vapor concentration. At the same time, the apparent reflectance information of satellite remote sensing data in the visible and near-infrared bands is extracted.

[0132] By fusing ground-based observation data with apparent reflectance information, and inverting the different response characteristics of the two to atmospheric conditions, spatially continuous atmospheric aerosol optical thickness and atmospheric water vapor content are obtained. Ground-based observation data provides precise atmospheric component concentrations at specific points, while apparent reflectance information reflects the scattering and absorption effects of light on a large scale. Combining the response differences between the two fills the spatial gaps in ground-based observations and generates key atmospheric parameters that cover the entire target area and are spatially continuous.

[0133] Furthermore, the surface emissivity or surface reflectivity in the corresponding band is derived from the localized relationship established by the historical clean day data of the target area. The radiance of the blackbody in the band corresponding to the surface temperature is derived from the actual surface temperature combined with the physical meaning of Planck's law. The observed zenith angle is determined by calculating the geometric relationship between the sensor observation position and the surface point of the target area. By decomposing the total radiance received by the sensor, the effective part of the surface radiation after atmospheric transmission and the radiation part generated by the atmosphere itself are separated to obtain the atmospheric transmittance and path radiation intensity, respectively.

[0134] The above calculations accurately distinguish the proportions of surface radiation and atmospheric interference in the sensor-received signals, providing a quantitative basis for two-way correction.

[0135] In general, the higher the atmospheric aerosol optical thickness and atmospheric water vapor content, the lower the atmospheric transmittance, the higher the path radiation intensity, the greater the surface emissivity or reflectivity, the higher the surface temperature, the greater the total radiance received by the sensor, the larger the observation zenith angle, the lower the atmospheric transmittance and the higher the path radiation intensity.

[0136] Furthermore, the three-dimensional thermal radiation field is attenuated and compensated based on atmospheric transmittance to obtain the corrected thermal radiation field. The thermal radiation value of each surface location in the three-dimensional thermal radiation field is divided by the atmospheric transmittance of the corresponding path to offset the absorption, scattering and attenuation effect of the atmosphere on the thermal infrared band radiation, ensuring that the corrected thermal radiation field can truly reflect the actual thermal radiation state of the surface.

[0137] Finally, background removal is performed on the video light signal based on the path radiation intensity to obtain the corrected light signal. The atmospheric path radiation intensity of the corresponding path is directly subtracted from the video light signal to remove background interference caused by atmospheric radiation itself and eliminate the attenuation effect of the atmosphere on the visible / near-infrared light signal, so that the corrected light signal accurately corresponds to the true reflected light intensity of the ground object.

[0138] It should be emphasized that those skilled in the art can also perform corrections through single-parameter calibration or based on experience. The present invention provides a preferred embodiment, which generates spatially continuous and accurate key atmospheric parameters by fusing and inverting ground-based monitoring data with satellite apparent reflectance and combining the different response characteristics of the two to atmospheric conditions, thus solving the problem of spatial heterogeneity in single-point monitoring extrapolation.

[0139] Overall, this scheme is based on the physical relationship of atmospheric radiation transmission, and calculates the atmospheric transmittance of the thermal infrared band and the path radiation intensity of the visible / near infrared band respectively, so as to achieve attenuation compensation of the three-dimensional thermal radiation field and background removal of video optical signals.

[0140] Step S50: Spatially overlay and analyze the trend of change of the corrected thermal radiation anomaly area and the video fire and smoke feature area, and determine whether it is a real fire based on the analysis and evaluation results. If it is determined to be a real fire, generate an early warning.

[0141] In a preferred embodiment, such as Figure 4 As shown, step S50 includes the following sub-steps:

[0142] Step S501: Extract the thermal radiation anomaly region from the corrected thermal radiation field;

[0143] Step S502: Extract the video smoke and fire feature region from the corrected optical signal;

[0144] Step S503: Calculate the overlapping area of ​​the thermal radiation anomaly area and the video fire and smoke feature area in the three-dimensional space scene;

[0145] Step S504: Construct the time-varying sequence of radiation intensity in the thermal radiation anomaly zone and the time-varying sequence of brightness in the video fire and smoke feature zone;

[0146] Step S505: Analyze the spatial stability of the overlapping region;

[0147] Step S506: Compare the trends of the time-varying sequences of radiation intensity and brightness.

[0148] Step S507: Combining the trend assessment results and spatial stability, determine the overall confidence level of the fire occurrence;

[0149] Step S508: Determine whether it is a real fire based on the overall confidence level. If it is determined to be a real fire, generate an early warning.

[0150] Specifically, the method of determining whether a fire is real based on the overall confidence level is as follows: the overall confidence level is compared with the judgment benchmark dynamically determined based on the current real-time weather conditions. If the overall confidence level exceeds the judgment benchmark dynamically determined based on the current real-time weather conditions, it is determined to be a real fire and an early warning is triggered.

[0151] In practice, the corrected thermal radiation field is traversed to screen out the surface locations where the radiation intensity exceeds the normal ground reference value. Adjacent areas with radiation values ​​exceeding the standard are integrated into spatially connected thermal radiation anomaly areas. At the same time, the corrected light signal is analyzed frame by frame to capture areas where the gray value is gradually distributed and the range continues to extend outward. Combined with the visual morphological characteristics of fire and smoke, video fire and smoke feature areas with diffusion characteristics are determined.

[0152] Furthermore, using the triangular facets of the three-dimensional scene as a spatial reference, the range of the triangular facets corresponding to the thermal radiation anomaly area and the video fire and smoke feature area are marked respectively. The triangular facet sets of the two regions are compared one by one, and the triangular facets that belong to both regions are summarized to form the overlapping area of ​​the two in three-dimensional space, thus clarifying the spatial location and coverage of the overlapping part.

[0153] Furthermore, at fixed time intervals within the video segment, the average radiation intensity of all triangular patches within the thermal radiation anomaly area is extracted sequentially. The values ​​at each time point are recorded and arranged in chronological order to form a radiation intensity time variation sequence. Simultaneously, for each time interval, the average brightness value of the video fire and smoke feature area is extracted and also arranged in chronological order to generate a brightness time variation sequence.

[0154] Furthermore, by observing the changes in the location and extent of the overlapping area during the video period, the spatial stability is high if there is no significant shift and the coverage area fluctuates little. By comparing the trends of the two time series, the morphological consistency is strong if the radiation intensity and brightness increase synchronously and the rhythm of change is consistent. Combining the assessment results of spatial stability and morphological consistency, the overall confidence level of the fire occurrence is obtained.

[0155] Finally, the judgment criteria are adjusted based on real-time meteorological data. By associating historical fire cases under different wind speed, wind direction and humidity conditions, the reasonable threshold for fire judgment under each meteorological scenario is clarified. When the comprehensive confidence level exceeds the dynamic threshold, it is directly judged as a real fire, triggering an early warning signal that includes the precise location, coverage and development trend of the fire.

[0156] It should be noted that those skilled in the art may also extract only one feature from the abnormal thermal radiation area or the video fire and smoke feature area. When the intensity of a single feature exceeds a fixed threshold, it can be directly judged as a fire and an early warning can be generated without spatial overlay and trend analysis; or a uniform and unchanging fire judgment benchmark can be set without dynamic adjustment based on real-time meteorological conditions. As long as the spatial overlap rate and feature intensity reach a fixed threshold, an early warning will be triggered.

[0157] This solution is a preferred implementation method. By accurately overlaying a three-dimensional spatial scene, it locks the spatially overlapping area of ​​thermal radiation and fire and smoke characteristics, and eliminates non-homogeneous interference signals. At the same time, it constructs a time change sequence of dual features, analyzes the trend synchronization, and ensures that the determination is of a continuously developing fire.

[0158] In summary, this solution dynamically adjusts the judgment criteria based on real-time meteorological conditions, adapts to the differences in fire characteristics under different environments, and achieves full-dimensional verification from space to time, significantly improving the accuracy and reliability of fire judgment and providing precise basis for fire response.

[0159] refer to Figure 5 The present invention also provides a comprehensive fire monitoring and early warning platform that integrates satellite remote sensing and video detection.

[0160] like Figure 5 As shown, this integrated fire monitoring and early warning platform combining satellite remote sensing and video detection includes:

[0161] The satellite data preprocessing module 501 is used to perform radiometric calibration on the satellite remote sensing data of the target area and generate a two-dimensional thermal radiation map with geographic coordinates.

[0162] The video 3D reconstruction module 502 is used to acquire multi-angle video surveillance data of the target monitoring area, perform 3D scene reconstruction based on the multi-angle video surveillance data, and generate a 3D spatial scene including terrain and equipment location.

[0163] The spatiotemporal fusion simulation module 503 is used to map the two-dimensional thermal radiation map to the corresponding surface location in the three-dimensional spatial scene, and to perform spatiotemporal simulation based on the time difference between the video and the satellite and the fire spread pattern, generating a three-dimensional thermal radiation field synchronized with the video time.

[0164] The multi-source data correction module 504 is used to perform bidirectional atmospheric attenuation correction on the three-dimensional thermal radiation field and video light signal using real-time meteorological data.

[0165] The intelligent analysis and early warning module 505 is used to spatially overlay and analyze the changing trends of the corrected thermal radiation anomaly area and the video fire and smoke feature area, and to determine whether it is a real fire based on the analysis and evaluation results. If it is determined to be a real fire, an early warning is generated.

[0166] Specifically, the specific operational procedures between the modules in the integrated fire monitoring and early warning platform that combines satellite remote sensing and video detection can be found in the integrated fire monitoring and early warning method that combines satellite remote sensing and video detection described above, and will not be repeated here.

[0167] In practice, the server-side equipment deployed in an integrated fire monitoring and early warning platform combining satellite remote sensing and video detection may consist of one or more devices. This integrated fire monitoring and early warning platform can be implemented as a business instance, a virtual machine, or hardware devices. For example, it can be implemented as a business instance deployed on one or more devices in a cloud node. Simply put, it can be understood as software deployed on a cloud node, providing integrated fire monitoring and early warning services to various user terminals. Alternatively, it can be implemented as a virtual machine deployed on one or more devices in a cloud node, with application software installed to manage various user terminals. Or, it can also be implemented as a server composed of numerous identical or different types of hardware devices, with one or more hardware devices configured to provide integrated fire monitoring and early warning services to various user terminals.

[0168] In terms of implementation, the integrated fire monitoring and early warning platform combining satellite remote sensing and video detection is mutually compatible with the user terminal. Specifically, if the integrated fire monitoring and early warning platform is implemented as an application installed on a cloud service platform, the user terminal acts as a client establishing a communication connection with that application; or if the integrated fire monitoring and early warning platform is implemented as a website, the user terminal acts as a webpage; or if the integrated fire monitoring and early warning platform is implemented as a cloud service platform, the user terminal acts as a mini-program within an instant messaging application.

[0169] This invention provides an integrated fire monitoring and early warning platform combining satellite remote sensing and video detection. This platform can be hosted on a cloud server. In terms of implementation, it can function as one or more service devices, or as an application installed in the cloud (e.g., a mobile service operator's server, server cluster, etc.), or it can be developed into a website. Depending on the functions implemented, the integrated fire monitoring and early warning platform can include a satellite data preprocessing module, a video 3D reconstruction module, a spatiotemporal fusion and simulation module, a multi-source data correction module, and an intelligent analysis and decision-making module. The modules in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0170] In this embodiment of the invention, in the integrated fire monitoring and early warning platform combining satellite remote sensing and video detection, each of the above-mentioned modules can be implemented independently and can call other modules. Here, "calling" can be understood as one module connecting to multiple modules of another type and providing corresponding services to those connected modules. In the integrated fire monitoring and early warning platform combining satellite remote sensing and video detection provided by this embodiment of the invention, the applicable scope of the platform architecture can be adjusted by adding modules and directly calling them without modifying the program code, achieving cluster-style horizontal expansion to quickly and flexibly expand the integrated fire monitoring and early warning platform. In practical applications, the above modules can be set in the same device or different devices, or they can be set in virtual devices, such as service instances in a cloud server.

[0171] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0172] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0173] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0174] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They do not limit the scope of protection of the present invention. All equivalent changes and modifications made within the scope of the claims of the present invention should fall within the scope of the claims of the present invention.

Claims

1. A comprehensive fire monitoring and early warning method integrating satellite remote sensing and video detection, characterized in that, include: Radiometric calibration is performed on satellite remote sensing data of the target area to generate a two-dimensional thermal radiation map with geographic coordinates, including: Based on the atmospheric-surface radiation coupling relationship, atmospheric correction is performed on the satellite remote sensing data to obtain the apparent radiation value at the top of the atmosphere; Identify cloud interference regions in the apparent radiation value of the top of the atmosphere, estimate the theoretical radiation of the ground surface under the clouds based on the surface thermal inertia and satellite transit time, replace the cloud interference regions, and obtain the cloud-corrected thermal radiation value. The corrected thermal radiation value of the cloud area is combined with the real-time acquired aerosol optical thickness data to perform aerosol scattering effect compensation and remove the specular reflection pseudo-high temperature points caused by solar flare angle to obtain the two-dimensional thermal radiation map. The atmospheric-surface radiation coupling relationship was obtained through the following method: Based on the analysis of historical clean day data observed over a long period in the target area, the localized synergistic variation relationship between surface emissivity and surface temperature was obtained. The real-time meteorological data acquired synchronously during satellite transit is analyzed to extract the water vapor content in the atmosphere of the target area. The localized coordinated change relationship and the water vapor content are substituted into the physical process describing thermal radiation transfer, and combined with satellite observation geometric parameters, the corresponding atmospheric path radiation and atmospheric transmittance are simulated and calculated to determine the atmospheric-surface radiation coupling relationship. Atmospheric path radiation and atmospheric transmittance can be realized using an atmospheric radiative transfer model, specifically: ; In the formula, Indicates atmospheric transmittance in the thermal infrared band. Indicates the first A type of atmospheric absorption component, where P is atmospheric pressure and T is temperature. Indicates the first The column content of various atmospheric absorption components in the vertical direction. This represents the zenith angle observed by the satellite. Represents the cosecant value of the zenith angle; Acquire multi-angle video surveillance data of the target monitoring area, and perform three-dimensional scene reconstruction based on the multi-angle video surveillance data to generate a three-dimensional spatial scene including terrain and equipment location, including: From the multi-angle video surveillance data, extract a set of visual feature points containing terrain edges and the outlines of fixed facilities within the target area; Based on synchronously acquired lidar point cloud data, spatial geometric correction is performed on the visual feature point set to obtain a sparse three-dimensional point cloud. Based on the constraint of terrain continuity, spatial interpolation is performed on the sparse 3D point cloud to obtain a dense 3D point cloud; The dense 3D point cloud is triangularly meshed to construct a surface, and vegetation texture information extracted from the multi-angle video surveillance data is fused to generate the 3D spatial scene. The two-dimensional thermal radiation map is mapped to the corresponding surface location of the three-dimensional spatial scene, and spatiotemporal extrapolation is performed based on the time difference between the video and the satellite and the fire spread pattern to generate a three-dimensional thermal radiation field synchronized with the video time. In the spatiotemporal simulation, the expansion of the abnormal high-temperature zone boundary and the change in radiation intensity are simulated based on the following fire spread physics model: ； In the formula, Indicates the moment in the video. The derived spatial position vector, Indicates the time of satellite transit The identified initial spatial position vector, Indicates the base spread rate, Indicates the wind acceleration factor. Represents the wind speed vector. This represents the angle between the wind direction vector and the direction vector of the fire spread normal. Represents the integral variable; The three-dimensional thermal radiation field and video light signal are subjected to bidirectional atmospheric attenuation correction using real-time meteorological data. In the process of performing two-way atmospheric attenuation correction, atmospheric transmittance and path radiation intensity are calculated according to the following model: ； In the formula, Indicates the satellite or video sensor in the band Total radiance received, Indicates the Earth's surface in the band The surface emissivity or surface reflectivity, This represents the temperature of a blackbody at the Earth's surface calculated using Planck's law. band radiance, Indicates along the observed zenith angle Atmospheric transmittance of the path, Indicates along the observed zenith angle Atmospheric path radiation along the path; The corrected thermal radiation anomaly area and the video fire and smoke feature area are spatially overlaid and their changing trends are analyzed. Based on the analysis and evaluation results, it is determined whether it is a real fire. If it is determined to be a real fire, an early warning is generated.

2. The integrated fire monitoring and early warning method combining satellite remote sensing and video detection according to claim 1, characterized in that, The process of mapping the two-dimensional thermal radiation map to the corresponding surface location in the three-dimensional spatial scene, and performing spatiotemporal extrapolation based on the time difference between the video and the satellite image and the fire spread pattern to generate a three-dimensional thermal radiation field synchronized with the video time includes: Based on the geometry of the surface triangular facets of the three-dimensional spatial scene and the geometry of satellite imaging, a projection relationship from the pixels of the thermal radiation map to the three-dimensional surface facets is established. The two-dimensional thermal radiation map is corrected pixel by pixel using the projection relationship to determine the target triangular facet corresponding to the radiation value of each pixel. Based on the geometric deformation ratio between the target triangular facet in three-dimensional space and the actual pixel projection, the radiation value is normalized by area to obtain a set of thermal radiation values ​​mapped to the surface location of the three-dimensional space scene. Based on the set of thermal radiation values ​​mapped to the surface location of the three-dimensional spatial scene, the abnormal high temperature zone is determined; Based on the synchronously acquired wind speed and direction data and time difference, the boundary expansion and radiation intensity change of the abnormal high temperature zone from satellite time to video time are deduced, and the deduced fire state at the video time is obtained. The simulated fire scene state at the video moment is assigned to the corresponding position in the three-dimensional spatial scene to form the three-dimensional thermal radiation field synchronized with the video time.

3. The integrated fire monitoring and early warning method combining satellite remote sensing and video detection according to claim 1, characterized in that, The bidirectional atmospheric attenuation correction of the three-dimensional thermal radiation field and video optical signal using real-time meteorological data includes: Key atmospheric parameters are determined based on the real-time meteorological data; these key atmospheric parameters include atmospheric aerosol optical thickness and atmospheric water vapor content. Based on the atmospheric radiative transfer relationship and the aforementioned key atmospheric parameters, atmospheric transmittance and path radiative intensity are calculated respectively. The three-dimensional thermal radiation field is attenuated and compensated according to the atmospheric transmittance to obtain the corrected thermal radiation field. Background removal is performed on the video optical signal based on the path radiation intensity to obtain the corrected optical signal.

4. The integrated fire monitoring and early warning method combining satellite remote sensing and video detection according to claim 3, characterized in that, The determination of key atmospheric parameters based on the real-time meteorological data includes: The observation data uploaded in real time by the ground-based monitoring network deployed in the target area is acquired; the observation data includes: aerosol concentration and atmospheric water vapor concentration; Simultaneously extract the apparent reflectance information of the satellite remote sensing data in the visible and near-infrared bands; The observation data and the apparent reflectance information are fused together, and the atmospheric aerosol optical thickness and the atmospheric water vapor content are obtained by coupling the different response characteristics of the two to atmospheric conditions.

5. The integrated fire monitoring and early warning method combining satellite remote sensing and video detection according to claim 1, characterized in that, The step of spatially overlaying and analyzing the trend of change between the corrected thermal radiation anomaly area and the video smoke feature area includes: Extract the thermal radiation anomaly region from the corrected thermal radiation field; Extract the video smoke and fire feature region from the corrected optical signal; Calculate the overlap area between the thermal radiation anomaly area and the video fire and smoke feature area in the three-dimensional spatial scene; Construct a time-varying sequence of radiation intensity in the thermal radiation anomaly zone and a time-varying sequence of brightness in the video fire and smoke feature zone.

6. The integrated fire monitoring and early warning method combining satellite remote sensing and video detection according to claim 5, characterized in that, The process of determining whether a fire is a real fire based on analysis and evaluation results, and generating an early warning if it is determined to be a real fire, includes: Analyze the spatial stability of the overlapping region; Compare the trends of the time-varying radiant intensity sequence and the time-varying luminance sequence; Based on the trend assessment results and the aforementioned spatial stability, the overall confidence level for the occurrence of the fire was determined; Based on the comprehensive confidence level, it is determined whether it is a real fire. If it is determined to be a real fire, an early warning is generated.

7. A comprehensive fire monitoring and early warning platform integrating satellite remote sensing and video detection, characterized in that, include: The satellite data preprocessing module is used to perform radiometric calibration on satellite remote sensing data of the target area and generate a two-dimensional thermal radiation map with geographic coordinates, including: Based on the atmospheric-surface radiation coupling relationship, atmospheric correction is performed on the satellite remote sensing data to obtain the apparent radiation value at the top of the atmosphere; Identify cloud interference regions in the apparent radiation value of the top of the atmosphere, estimate the theoretical radiation of the ground surface under the clouds based on the surface thermal inertia and satellite transit time, replace the cloud interference regions, and obtain the cloud-corrected thermal radiation value. The corrected thermal radiation value of the cloud area is combined with the real-time acquired aerosol optical thickness data to perform aerosol scattering effect compensation and remove the specular reflection pseudo-high temperature points caused by solar flare angle to obtain the two-dimensional thermal radiation map. The atmospheric-surface radiation coupling relationship was obtained through the following method: Based on the analysis of historical clean day data observed over a long period in the target area, the localized synergistic variation relationship between surface emissivity and surface temperature was obtained. The real-time meteorological data acquired synchronously during satellite transit is analyzed to extract the water vapor content in the atmosphere of the target area. The localized coordinated change relationship and the water vapor content are substituted into the physical process describing thermal radiation transfer, and combined with satellite observation geometric parameters, the corresponding atmospheric path radiation and atmospheric transmittance are simulated and calculated to determine the atmospheric-surface radiation coupling relationship. Atmospheric path radiation and atmospheric transmittance can be realized using an atmospheric radiative transfer model, specifically: ; In the formula, Indicates atmospheric transmittance in the thermal infrared band. Indicates the first A type of atmospheric absorption component, where P is atmospheric pressure and T is temperature. Indicates the first The column content of various atmospheric absorption components in the vertical direction. This represents the zenith angle observed by the satellite. Represents the cosecant value of the zenith angle; The video 3D reconstruction module is used to acquire multi-angle video surveillance data of the target monitoring area, and perform 3D scene reconstruction based on the multi-angle video surveillance data to generate a 3D spatial scene including terrain and equipment location, including: From the multi-angle video surveillance data, extract a set of visual feature points containing terrain edges and the outlines of fixed facilities within the target area; Based on synchronously acquired lidar point cloud data, spatial geometric correction is performed on the visual feature point set to obtain a sparse three-dimensional point cloud. Based on the constraint of terrain continuity, spatial interpolation is performed on the sparse 3D point cloud to obtain a dense 3D point cloud; The dense 3D point cloud is triangularly meshed to construct a surface, and vegetation texture information extracted from the multi-angle video surveillance data is fused to generate the 3D spatial scene. The spatiotemporal fusion simulation module is used to map the two-dimensional thermal radiation map to the corresponding surface location of the three-dimensional spatial scene, and to perform spatiotemporal simulation based on the time difference between the video and the satellite and the fire spread pattern to generate a three-dimensional thermal radiation field synchronized with the video time. In the spatiotemporal simulation, the expansion of the abnormal high-temperature zone boundary and the change in radiation intensity are simulated based on the following fire spread physics model: ； In the formula, Indicates the moment in the video. The derived spatial position vector, Indicates the time of satellite transit The identified initial spatial position vector, Indicates the base spread rate, Indicates the wind acceleration factor. Represents the wind speed vector. This represents the angle between the wind direction vector and the direction vector of the fire spread normal. Represents the integral variable; A multi-source data correction module is used to perform bidirectional atmospheric attenuation correction on the three-dimensional thermal radiation field and video light signal using real-time meteorological data. In the process of performing two-way atmospheric attenuation correction, atmospheric transmittance and path radiation intensity are calculated according to the following model: ； In the formula, Indicates the satellite or video sensor in the band Total radiance received, Indicates the Earth's surface in the band The surface emissivity or surface reflectivity, This represents the temperature of a blackbody at the Earth's surface calculated using Planck's law. band radiance, Indicates along the observed zenith angle Atmospheric transmittance of the path, Indicates along the observed zenith angle Atmospheric path radiation along the path; The intelligent analysis and early warning module is used to spatially overlay and analyze the changing trends of the corrected thermal radiation anomaly area and the video fire and smoke feature area, and to determine whether it is a real fire based on the analysis and evaluation results. If it is determined to be a real fire, an early warning is generated.

Images