A polar-orbit satellite-assisted identification method and device for suspected fire points of geostationary satellites

Abstract

The present invention discloses a polar-orbit satellite-assisted identification method and device for geostationary satellite suspected fire points, the method comprising: extracting at least one suspected fire point based on geostationary satellite remote sensing images; extracting geostationary satellite remote sensing images and polar-orbit satellite remote sensing images within a preset range of the location of at least one of the suspected fire points and a certain time period before the occurrence time, and at the same time extracting surface characteristic parameters within a preset range of the location of at least one of the suspected fire points, generating time series multi-channel fused raster data; inputting the generated time series multi-channel fused raster data into a trained multi-layer neural network model, and outputting the identification result of at least one of the suspected fire points. The present invention extracts geostationary satellite images of a period of time before the suspected fire point and the surrounding area, and superimposes polar-orbit satellite images, combines surface characteristic parameters, and uses a multi-layer neural network to construct a model to identify and confirm the fire point, thereby improving the accuracy and computational efficiency of fire point identification.

Description

Technical Field

[0001] The present invention relates to remote sensing fire point monitoring, and in particular to a polar-orbit satellite-assisted identification method and device for a geostationary satellite suspected fire point. Background Art

[0002] Fire incidents cause serious losses to human life and property. Fire point monitoring is of great significance in helping humans control the combustion process. Satellite remote sensing fire point identification is an important means of fire point monitoring. Among them, geostationary meteorological satellites have the advantage of good real-time performance in fire point monitoring, but have low spatial resolution and high false alarm rate of fire point identification.

[0003] There are related technologies that use fire products identified by polar-orbiting meteorological satellites to verify fires identified by geostationary satellites. However, when polar-orbiting meteorological satellites generate fire products, they only use their own data, and the accuracy of the fire data obtained is not high, and the ability to assist geostationary satellites in identification is limited. In addition, there are related technologies that identify fires by fusing multi-source satellite remote sensing data for joint processing, but the calculation efficiency is low. Summary of the invention

[0004] In order to overcome the deficiencies of the above-mentioned prior art, the present invention provides a polar-orbiting satellite-assisted identification method and device for suspected fire points of geostationary satellites, which extracts geostationary satellite images of a period of time before the suspected fire point and the surrounding area and superimposes polar-orbiting satellite images, combines surface feature parameters, and uses a multi-layer neural network to build a model to identify and confirm the fire point, thereby improving the fire point identification accuracy and computational efficiency.

[0005] According to one aspect of the present invention, a polar-orbiting satellite-assisted identification method for suspected fire points of a geostationary satellite is provided, comprising:

[0006] Extract at least one suspected fire point based on geostationary satellite remote sensing images;

[0007] Extracting a geostationary satellite remote sensing image and a polar-orbiting satellite remote sensing image of a preset range of at least one suspected fire point and a certain time before the occurrence time, and at the same time extracting a surface characteristic parameter of a preset range of at least one suspected fire point, to generate time series multi-channel fused raster data;

[0008] The generated time series multi-channel fusion raster data is input into the trained multi-layer neural network model to output at least one discrimination result of the suspected fire point; wherein the training of the multi-layer neural network model includes:

[0009] Obtain remote sensing image data and surface characteristic parameters from geostationary and polar-orbiting satellites in the study area, and also obtain fire point data in the study area;

[0010] Extract multiple suspected fire points based on the remote sensing image data of the geostationary satellite, and extract the geostationary satellite remote sensing image and polar-orbiting satellite remote sensing image of a preset range of each suspected fire point and a certain time before the occurrence time, as well as the surface characteristic parameters of the preset range of each suspected fire point, to generate time series multi-channel fused raster data of each suspected fire point;

[0011] The time series multi-channel fused raster data of each suspected fire point are matched with the fire point data in time to obtain a modeling sample set;

[0012] A multi-layer neural network model is constructed, and the multi-layer neural network model is trained using the modeling sample set to obtain a trained multi-layer neural network model.

[0013] As a further technical solution, generating time series multi-channel fused raster data also includes:

[0014] Obtain geostationary satellite remote sensing images within a set time series and polar-orbiting satellite remote sensing images within the same time range, and extract geostationary satellite image data and polar-orbiting satellite image data within the study area;

[0015] Obtain the surface characteristic parameters of the same study area;

[0016] On the geostationary satellite remote sensing image with suspected fire points, polar-orbiting satellite remote sensing images of a preset range of the location of each suspected fire point and a certain time period before the occurrence time and surface characteristic parameters of a preset range of the location of each suspected fire point are superimposed to form time-series multi-channel fused raster data of each suspected fire point.

[0017] As a further technical solution, the surface characteristic parameters include: surface temperature, surface humidity, vegetation coverage and terrain characteristics.

[0018] As a further technical solution, after obtaining the geostationary satellite image data within the study area, it also includes: extracting suspected fire points based on the geostationary satellite remote sensing image data.

[0019] As a further technical solution, a multi-layer neural network model is constructed, and the multi-layer neural network model is trained using the modeling sample set, including:

[0020] A multi-layer neural network model is constructed based on the back propagation neural network. At the suspected fire point, the time series multi-channel fused raster data corresponding to the suspected fire point is used as the input of the multi-layer neural network model, and the identification of the fire point / non-fire point at the corresponding time is used as the output of the multi-layer neural network model.

[0021] The multi-layer neural network model is trained and verified using the modeling sample set, and the model parameters are saved.

[0022] According to one aspect of the present invention, a polar-orbit satellite-assisted identification device for suspected fire points of a geostationary satellite is provided, comprising:

[0023] The first main module is used to extract at least one suspected fire point based on the geostationary satellite remote sensing image;

[0024] The second main module is used to extract the geostationary satellite remote sensing image and polar-orbiting satellite remote sensing image of at least one of the preset ranges of the suspected fire point and a certain length of time before the occurrence time, and at the same time extract the surface characteristic parameters of at least one of the preset ranges of the suspected fire point to generate time series multi-channel fused raster data;

[0025] The third main module is used to input the generated time series multi-channel fusion raster data into the trained multi-layer neural network model, and output the discrimination result of at least one suspected fire point; wherein the training of the multi-layer neural network model includes:

[0026] Obtain remote sensing image data and surface characteristic parameters from geostationary and polar-orbiting satellites in the study area, and also obtain fire point data in the study area;

[0027] Extract multiple suspected fire points based on the remote sensing image data of the geostationary satellite, and extract the geostationary satellite remote sensing image and polar-orbiting satellite remote sensing image of a preset range of each suspected fire point and a certain time before the occurrence time, as well as the surface characteristic parameters of the preset range of each suspected fire point, to generate time series multi-channel fused raster data of each suspected fire point;

[0028] The time series multi-channel fused raster data of each suspected fire point are matched with the fire point data in time to obtain a modeling sample set;

[0029] A multi-layer neural network model is constructed, and the multi-layer neural network model is trained using the modeling sample set to obtain a trained multi-layer neural network model.

[0030] According to one aspect of the present invention, there is provided a computing device, including a processor and a memory, wherein the memory stores program instructions executed by the processor, and the processor calls the program instructions to execute the steps of a polar-orbiting satellite-assisted identification method for suspected fire points of a geostationary satellite.

[0031] According to one aspect of the present specification, there is provided a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions enable the computer to execute the steps of a polar-orbiting satellite-assisted identification method for suspected fire points of a geostationary satellite.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] The present invention extracts the geostationary satellite images and polar-orbit satellite images of the suspected fire point before and around the suspected fire point, combines the surface characteristic parameters, and uses a multi-layer neural network to build a model to identify and confirm the fire point. Compared with the method of verifying the fire point product using the polar-orbit satellite, the accuracy of fire point identification can be improved and the misjudgment rate can be reduced; compared with the joint processing method of multi-source satellite remote sensing data, the calculation efficiency can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, a brief introduction is given below to the drawings used in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0035] Figure 1 A schematic flow chart of a polar-orbiting satellite-assisted identification method for suspected fire points on a geostationary satellite provided in an embodiment of the present invention.

[0036] Figure 2 A schematic diagram of the training process of a multi-layer neural network model provided in an embodiment of the present invention.

[0037] Figure 3 A schematic structural diagram of a polar-orbiting satellite-assisted identification device for suspected fire points on a geostationary satellite provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] It should be noted that:

[0039] Unless otherwise specified, the technologies involved in the specification of the present invention are all well-known technologies.

[0040] The geostationary satellite and geostationary meteorological satellite mentioned in the specification of the present invention refer to the same thing.

[0041] The terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions, for example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to the steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products or apparatuses.

[0042] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In addition, the technical features in the various embodiments or single embodiments provided by the present invention are arbitrarily combined with each other to form a new technical solution. This combination is not restricted by the sequence of steps and / or the structural composition mode, but must be based on the ability of ordinary technicians in this field to achieve. When the combination of technical solutions is contradictory or cannot be achieved, it should be considered that this combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0043] The embodiment of the present invention provides a polar-orbiting satellite-assisted identification method for suspected fire points of a geostationary satellite, such as Figure 1 As shown, including:

[0044] Extract at least one suspected fire point based on geostationary satellite remote sensing images;

[0045] Extracting a geostationary satellite remote sensing image and a polar-orbiting satellite remote sensing image of a preset range of at least one suspected fire point and a certain time before the occurrence time, and at the same time extracting a surface characteristic parameter of a preset range of at least one suspected fire point, to generate time series multi-channel fused raster data;

[0046] The generated time series multi-channel fused raster data is input into the trained multi-layer neural network model to output a discrimination result of at least one of the suspected fire points.

[0047] The embodiment of the present invention extracts the superposition of stationary satellite images and polar-orbiting satellite images of the period before the suspected fire point and the surrounding area, combines surface feature parameters, and uses a multi-layer neural network model to identify and confirm the fire point, thereby improving both the accuracy of fire point identification and the calculation efficiency.

[0048] In the embodiment of the present invention, Figure 2 As shown, the training of the multi-layer neural network model includes:

[0049] S1, download the remote sensing images of geostationary meteorological satellites within a certain time series and polar-orbiting satellite remote sensing images within the same time range, and extract the image data within the study area.

[0050] S2, download the fire point database within the same time series and extract the fire point data within the area.

[0051] S3, download the surface characteristic parameters of the study area (surface temperature, surface humidity, vegetation coverage, terrain characteristics).

[0052] S4, using the geostationary meteorological satellite remote sensing image to extract fire points as suspected fire points. The extracted suspected fire points may be one or more.

[0053] S5, extracts the geostationary satellite imagery and polar-orbiting satellite imagery data of a certain range (e.g., 10 km in the north-south direction, 10 km in the east-west direction) around the suspected fire point and a period of time (e.g., 12 hours) before the fire occurs, and uses the surface characteristic parameters as additional layer channels to generate time series multi-channel fused raster data.

[0054] S6, matching the time series multi-channel raster data with the fire point data to obtain a sample set for modeling, and dividing the data set into a training set and a validation set.

[0055] S7, construct a multi-layer neural network model. At the suspected fire point, the multi-channel fused raster data is used as the input layer, and the identification of the fire point / non-fire point at the corresponding time is used as the output layer.

[0056] S8, input the training set and the validation set into the multi-layer neural network model for training, and save the model parameters.

[0057] After the trained model is obtained, fire point monitoring is performed based on the multi-layer neural network model. At the suspected fire point, the multi-channel fused raster data is used as input, and the trained model is used to output the fire point discrimination result.

[0058] As a preferred embodiment, the embodiment of the present invention introduces an embodiment of fire point monitoring modeling in a certain area based on a back-propagation neural network using FY-4B satellite products and FY-3E satellite products.

[0059] (1) Download FY-4B satellite remote sensing images from October 2023 to October 2024 and FY-3E satellite remote sensing images of the same time range, and extract image data in the area.

[0060] (2) Download the fire point database within the same time series and extract the fire point data in the area.

[0061] (3) Download the surface characteristic parameters of the study area (surface temperature, surface humidity, vegetation coverage, and terrain characteristics).

[0062] (4) Use FY-4B satellite remote sensing images to extract fire spots as suspected fire spots.

[0063] (5) FY-4B and FY-3E satellite image data were extracted from the area 10 km north-south and 10 km east-west around the suspected fire point 12 hours before the fire occurred. At the same time, surface characteristic parameters (surface temperature, surface humidity, vegetation coverage, terrain characteristics) were used as additional layer channels to generate time series multi-channel fused raster data.

[0064] (6) Match the time series multi-channel raster data with the fire point data to obtain a sample set for modeling. The data set is divided into a training set and a validation set, with a ratio of 8:2.

[0065] (7) Construct a multi-layer neural network model. At the suspected fire point, the multi-channel fused raster data is used as the input layer, and the identification of the fire point / non-fire point at the corresponding time (marked as 1 and 0 respectively) is used as the output layer.

[0066] (8) Input the training set and validation set into the multi-layer neural network model for training and save the model parameters.

[0067] After the model training is completed, the fire point monitoring is carried out based on the multi-layer neural network model. At the suspected fire point, the multi-channel fused raster data is used as input to obtain the fire point identification result.

[0068] The implementation basis of each embodiment of the present invention is to implement programmed processing through a device with a processor function. Therefore, in engineering practice, the technical solutions and functions of each embodiment of the present invention are encapsulated into various modules. Based on this reality, on the basis of the above embodiments, an embodiment of the present invention provides a polar-orbiting satellite-assisted discrimination device for suspected fire points of a geostationary satellite, which is used to execute a polar-orbiting satellite-assisted discrimination method for suspected fire points of a geostationary satellite in the above method embodiment.

[0069] See also Figure 3The device includes: a first main module, which is used to extract at least one suspected fire point based on the geostationary satellite remote sensing image; a second main module, which is used to extract at least one geostationary satellite remote sensing image and a polar-orbiting satellite remote sensing image within a preset range of the location of the suspected fire point and a certain time before the occurrence time, and at the same time extract at least one surface characteristic parameter within a preset range of the location of the suspected fire point to generate time series multi-channel fused raster data; a third main module, which is used to input the generated time series multi-channel fused raster data into the trained multi-layer neural network model, and output the discrimination result of at least one suspected fire point; wherein the training of the multi-layer neural network model includes: obtaining remote sensing images of geostationary satellites and polar-orbiting satellites in the study area; The method comprises the following steps: obtaining the time series multi-channel fused raster data of each suspected fire point according to the remote sensing image data of the geostationary satellite and the surface characteristic parameters, and obtaining the fire point data in the study area at the same time; extracting multiple suspected fire points according to the remote sensing image data of the geostationary satellite, and extracting the geostationary satellite remote sensing image and the polar-orbiting satellite remote sensing image within a preset range of the location of each suspected fire point and a certain length of time before the occurrence time, as well as the surface characteristic parameters within a preset range of the location of each suspected fire point, to generate the time series multi-channel fused raster data of each suspected fire point; matching the time series multi-channel fused raster data of each suspected fire point with the fire point data according to time, to obtain a modeling sample set; constructing a multi-layer neural network model, and training the multi-layer neural network model using the modeling sample set, to obtain a trained multi-layer neural network model.

[0070] An embodiment of the present invention provides a polar-orbit satellite-assisted identification device for geostationary satellite suspected fire points, using Figure 3 Several modules in it extract the static satellite images of the suspected fire point and the surrounding area in the period before the fire point and the superposition of polar-orbiting satellite images, combine the surface feature parameters, and use a multi-layer neural network to build a model to identify and confirm the fire point, thereby improving the accuracy and calculation efficiency of fire point identification.

[0071] It should be noted that the device embodiments provided by the present invention are used to implement the methods in the above method embodiments as well as the methods in other method embodiments provided by the present invention. The only difference is that the corresponding functional modules are set, and the principle is basically the same as the principle of the above device embodiments provided by the present invention. As long as the technical personnel in the field refer to the specific technical solutions in other method embodiments on the basis of the above device embodiments, obtain the corresponding technical means and the technical solutions composed of these technical means by combining the technical features, and on the premise of ensuring the practicality of the technical solutions, improve the modules in the above device embodiments to obtain the corresponding device class embodiments, which are used to implement the methods in other method class embodiments. For example:

[0072] Based on the content of the above device embodiment, as a preferred embodiment, in a polar-orbiting satellite auxiliary identification device for suspected fire points of a geostationary satellite provided in an embodiment of the present invention, the second main module is further used to execute the following instructions:

[0073] Obtain geostationary satellite remote sensing images within a set time series and polar-orbiting satellite remote sensing images within the same time range, and extract geostationary satellite image data and polar-orbiting satellite image data within the study area;

[0074] Obtain the surface characteristic parameters of the same study area;

[0075] On the geostationary satellite remote sensing image with suspected fire points, polar-orbiting satellite remote sensing images of a preset range of the location of each suspected fire point and a certain time period before the occurrence time and surface characteristic parameters of a preset range of the location of each suspected fire point are superimposed to form time-series multi-channel fused raster data of each suspected fire point.

[0076] Based on the contents of the above-mentioned device embodiments, as a preferred embodiment, in a polar-orbiting satellite-assisted identification device for suspected fire points of a geostationary satellite provided in an embodiment of the present invention, the surface characteristic parameters include: surface temperature, surface humidity, vegetation coverage and terrain characteristics.

[0077] Based on the content of the above device embodiment, as a preferred embodiment, in a polar-orbiting satellite auxiliary identification device for suspected fire points of a geostationary satellite provided in an embodiment of the present invention, the second main module is further used to execute the following instructions:

[0078] Extract suspected fire points based on geostationary satellite remote sensing image data.

[0079] Based on the content of the above device embodiment, as a preferred embodiment, in a polar-orbiting satellite auxiliary identification device for suspected fire points of a geostationary satellite provided in an embodiment of the present invention, the third main module is further used to execute the following instructions:

[0080] A multi-layer neural network model is constructed based on the back propagation neural network. At the suspected fire point, the time series multi-channel fused raster data corresponding to the suspected fire point is used as the input of the multi-layer neural network model, and the identification of the fire point / non-fire point at the corresponding time is used as the output of the multi-layer neural network model.

[0081] The multi-layer neural network model is trained and verified using the modeling sample set, and the model parameters are saved.

[0082] Based on the same inventive concept as the above-mentioned embodiment, an embodiment of the present invention also provides a computing device, including a processor and a memory, wherein the memory stores program instructions executed by the processor, and the processor calls the program instructions to execute the steps of the polar-orbiting satellite-assisted identification method of a suspected fire point of a geostationary satellite.

[0083] When using the computing device provided by the embodiment of the present invention to identify suspected fire points, the fire point is identified and confirmed by extracting the static satellite images of the suspected fire point and the surrounding area for a period of time before the fire point and superimposing the polar-orbiting satellite images, combining the surface feature parameters, and using a multi-layer neural network to build a model, thereby improving the accuracy of fire point identification and the calculation efficiency.

[0084] The embodiment of the present invention further provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, wherein the computer instructions enable the computer to execute the steps of the polar-orbiting satellite-assisted identification method of a suspected fire point of a geostationary satellite, including:

[0085] Extract at least one suspected fire point based on geostationary satellite remote sensing images;

[0086] Extracting a geostationary satellite remote sensing image and a polar-orbiting satellite remote sensing image of a preset range of at least one suspected fire point and a certain time before the occurrence time, and at the same time extracting a surface characteristic parameter of a preset range of at least one suspected fire point, to generate time series multi-channel fused raster data;

[0087] The generated time series multi-channel fusion raster data is input into the trained multi-layer neural network model to output at least one discrimination result of the suspected fire point; wherein the training of the multi-layer neural network model includes:

[0088] Obtain remote sensing image data and surface characteristic parameters from geostationary and polar-orbiting satellites in the study area, and also obtain fire point data in the study area;

[0089] Extract multiple suspected fire points based on the remote sensing image data of the geostationary satellite, and extract the geostationary satellite remote sensing image and polar-orbiting satellite remote sensing image of a preset range of each suspected fire point and a certain time before the occurrence time, as well as the surface characteristic parameters of the preset range of each suspected fire point, to generate time series multi-channel fused raster data of each suspected fire point;

[0090] The time series multi-channel fused raster data of each suspected fire point are matched with the fire point data in time to obtain a modeling sample set;

[0091] A multi-layer neural network model is constructed, and the multi-layer neural network model is trained using the modeling sample set to obtain a trained multi-layer neural network model.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A polar-orbiting satellite-assisted identification method for suspected fire points of geostationary satellites, characterized in that: include: Extract at least one suspected fire point based on geostationary satellite remote sensing images; Extracting a geostationary satellite remote sensing image and a polar-orbiting satellite remote sensing image of a preset range of at least one suspected fire point and a certain time before the occurrence time, and at the same time extracting a surface characteristic parameter of a preset range of at least one suspected fire point, to generate time series multi-channel fused raster data; The generated time series multi-channel fusion raster data is input into the trained multi-layer neural network model to output at least one discrimination result of the suspected fire point; wherein the training of the multi-layer neural network model includes: Obtain remote sensing image data and surface characteristic parameters from geostationary and polar-orbiting satellites in the study area, and also obtain fire point data in the study area; Extract multiple suspected fire points based on the remote sensing image data of the geostationary satellite, and extract the geostationary satellite remote sensing image and polar-orbiting satellite remote sensing image of a preset range of each suspected fire point and a certain time before the occurrence time, as well as the surface characteristic parameters of the preset range of each suspected fire point, to generate time series multi-channel fused raster data of each suspected fire point; The time series multi-channel fused raster data of each suspected fire point are matched with the fire point data in time to obtain a modeling sample set; A multi-layer neural network model is constructed, and the multi-layer neural network model is trained using the modeling sample set to obtain a trained multi-layer neural network model.

2. According to claim 1, a polar-orbiting satellite-assisted identification method for suspected fire points of a geostationary satellite is characterized in that: Generate time series multi-channel fused raster data, including: Obtain geostationary satellite remote sensing images within a set time series and polar-orbiting satellite remote sensing images within the same time range, and extract geostationary satellite image data and polar-orbiting satellite image data within the study area; Obtain the surface characteristic parameters of the same study area; On the geostationary satellite remote sensing image with suspected fire points, polar-orbiting satellite remote sensing images of a preset range of the location of each suspected fire point and a certain time period before the occurrence time and surface characteristic parameters of a preset range of the location of each suspected fire point are superimposed to form time-series multi-channel fused raster data of each suspected fire point.

3. According to claim 2, a polar-orbiting satellite-assisted identification method for suspected fire points of a geostationary satellite is characterized in that: The surface characteristic parameters include: surface temperature, surface humidity, vegetation coverage and terrain characteristics.

4. According to claim 2, a polar-orbiting satellite-assisted identification method for suspected fire points of a geostationary satellite is characterized in that: After obtaining the geostationary satellite image data within the study area, the method also includes: extracting suspected fire points based on the geostationary satellite remote sensing image data.

5. According to claim 1, a polar-orbiting satellite-assisted identification method for suspected fire points of a geostationary satellite is characterized in that: Constructing a multi-layer neural network model and training the multi-layer neural network model using the modeling sample set, including: A multi-layer neural network model is constructed based on the back propagation neural network. At the suspected fire point, the time series multi-channel fused raster data corresponding to the suspected fire point is used as the input of the multi-layer neural network model, and the identification of the fire point / non-fire point at the corresponding time is used as the output of the multi-layer neural network model. The multi-layer neural network model is trained and verified using the modeling sample set, and the model parameters are saved.

6. A polar-orbit satellite-assisted identification device for suspected fire points of geostationary satellites, characterized in that: include: The first main module is used to extract at least one suspected fire point based on the geostationary satellite remote sensing image; The second main module is used to extract the geostationary satellite remote sensing image and polar-orbiting satellite remote sensing image of at least one of the preset ranges of the suspected fire point and a certain length of time before the occurrence time, and at the same time extract the surface characteristic parameters of at least one of the preset ranges of the suspected fire point to generate time series multi-channel fused raster data; The third main module is used to input the generated time series multi-channel fusion raster data into the trained multi-layer neural network model, and output the discrimination result of at least one suspected fire point; wherein the training of the multi-layer neural network model includes: Obtain remote sensing image data and surface characteristic parameters from geostationary and polar-orbiting satellites in the study area, and also obtain fire point data in the study area; Extract multiple suspected fire points based on the remote sensing image data of the geostationary satellite, and extract the geostationary satellite remote sensing image and polar-orbiting satellite remote sensing image of a preset range of each suspected fire point and a certain time before the occurrence time, as well as the surface characteristic parameters of the preset range of each suspected fire point, to generate time series multi-channel fused raster data of each suspected fire point; The time series multi-channel fused raster data of each suspected fire point are matched with the fire point data in time to obtain a modeling sample set; A multi-layer neural network model is constructed, and the multi-layer neural network model is trained using the modeling sample set to obtain a trained multi-layer neural network model.

7. A computing device, characterized in that The method comprises a processor and a memory, wherein the memory stores program instructions executed by the processor, and the processor calls the program instructions to execute the steps of a polar-orbiting satellite-assisted identification method for suspected fire points of a geostationary satellite as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions enable the computer to execute the steps of a polar-orbiting satellite-assisted identification method for suspected fire points of a geostationary satellite as described in any one of claims 1 to 5.

Images