Method and device for detecting haze distribution and aerosol optical thickness based on satellite remote sensing

Data is obtained and processed through satellite remote sensing technology, and artificial neural networks are used to classify haze distribution information, which solves the problem of difficult-to-reflect haze distribution caused by sparse ground monitoring sites, and achieves efficient monitoring widely used in pollutant monitoring and environmental analysis.

CN114418970BActive Publication Date: 2024-11-26AEROSPACE INFORMATION RES INST CAS
View PDF 8 Cites 0 Cited by

Patent Information

Application Number
CN202111673313.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-12-31
Publication Date
2024-11-26
Estimated Expiration
2041-12-31

AI Technical Summary

Technical Problem

The existing technology is difficult to fully reflect the spatial distribution of haze, and the ground environment monitoring stations are sparse, making it difficult to effectively monitor regional atmospheric environmental problems.

Method used

Using satellite remote sensing based on haze distribution and aerosol optical thickness detection methods, by obtaining satellite remote sensing data, inversion calculation and hyperspectral image data processing, artificial neural network is used to classify aerosol optical thickness and haze feature maps. Obtain information on haze distribution.

Benefits of technology

It provides haze distribution and aerosol optical thickness detection in a wide range, supports the applications of pollutant monitoring, pollution event determination, pollution source analysis and regional transportation, and improves the coverage and accuracy of atmospheric environmental monitoring.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114418970B_ABST
    Figure CN114418970B_ABST
Patent Text Reader

Abstract

The present invention relates to a method and device for detecting haze distribution and aerosol optical thickness based on satellite remote sensing, including obtaining satellite remote sensing data of a target monitoring area, performing inversion calculation on the satellite remote sensing data of each target area respectively, and determining the aerosol optical thickness of each target area; obtaining haze characteristic maps of each target area according to the hyperspectral image data of each target area, and spatially matching the aerosol optical thickness of each target area with the haze characteristic map of the target area; and classifying the aerosol optical thickness of each target area and the haze characteristic map of the target area by using an artificial neural network to obtain haze distribution information of the target monitoring area. The satellite remote sensing of the present invention can provide haze distribution and aerosol optical thickness detection in a wide range, and has a wide range of applications in pollutant monitoring, determination of pollution events, pollution source analysis, and regional transportation of pollutants.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of information processing technology, and in particular to a method and device for detecting haze distribution and aerosol optical thickness based on satellite remote sensing. Background Art

[0002] With the rapid development of the national economy, the scale of cities has continued to expand. Industrialization has led to rapid social and economic development, but it has also brought about a large consumption of resources and serious damage to the atmospheric environment.

[0003] Especially in recent years, regional atmospheric environmental problems have become increasingly prominent. Currently, ground environmental monitoring stations are mainly established to monitor particulate matter and pollutant gas concentrations, but these stations are often sparse and concentrated in cities, making it difficult to fully reflect the spatial distribution of haze. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a method and device for detecting haze distribution and aerosol optical thickness based on satellite remote sensing in view of the deficiencies in the prior art.

[0005] The technical solution of the present invention to solve the above technical problems is as follows:

[0006] A method for detecting haze distribution and aerosol optical thickness based on satellite remote sensing, comprising:

[0007] Acquire satellite remote sensing data of a target monitoring area, divide the target monitoring area into a plurality of target areas, perform inversion calculations on the satellite remote sensing data of each target area, and determine the aerosol optical thickness of each target area;

[0008] According to the satellite remote sensing data of each target area, the hyperspectral image data of each target area is obtained, and according to the hyperspectral image data of each target area, the haze characteristic map of each target area is obtained, and the aerosol optical thickness of each target area and the haze characteristic map of the target area are spatially matched;

[0009] The aerosol optical thickness of each target area and the haze characteristic map of the target area are classified by using an artificial neural network to obtain the haze distribution information of the target monitoring area.

[0010] Based on the above technical solution, the present invention can also be improved as follows.

[0011] Furthermore, obtaining the haze characteristic map of each target area according to the hyperspectral image data of each target area specifically includes:

[0012] The hyperspectral image data is smoothed and filtered, spectral line noise is removed, and then normalized to obtain pre-processed hyperspectral image data;

[0013] The pre-processed hyperspectral image data is subjected to a weighted grayscale algorithm to obtain a grayscale value of the hyperspectral image data;

[0014] A haze characteristic map of the target area is obtained according to the preprocessed hyperspectral image data and the grayscale value of the hyperspectral image.

[0015] Furthermore, the aerosol optical thickness of each target area and the haze characteristic map of the target area are classified by using an artificial neural network to obtain the haze distribution information of the target monitoring area, specifically including:

[0016] Extracting features of a preset type from the haze feature map, and constructing combined features corresponding to the haze feature map based on the extracted features;

[0017] The combined features corresponding to the haze feature map and the aerosol optical thickness of the target area are used as inputs of an artificial neural network model, and the haze distribution information of the target monitoring area is determined according to the output result of the artificial neural network model.

[0018] Furthermore, the artificial neural network model is established by:

[0019] The category labeling results corresponding to the historical haze characteristic maps at each historical time are used as training data;

[0020] Extracting a preset number of preset type features from the historical haze feature map, and obtaining a combination feature corresponding to the historical haze feature map based on the extracted type features;

[0021] Taking the combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map as inputs of a classification model, and obtaining output results of the classification model for the combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map;

[0022] Obtaining a corresponding loss function according to the output result of the classification model for the aerosol optical thickness corresponding to the historical time of the historical haze characteristic map and the historical time, and the category labeling result corresponding to the historical haze characteristic map at the historical time;

[0023] According to the loss function, the loss function of the classification model is determined, and the loss function of the classification model is used to adjust the parameters of the classification model to obtain the artificial neural network model.

[0024] Furthermore, the output result of the aerosol optical thickness corresponding to the historical time of the historical haze feature map and the historical time of the historical haze feature map according to the classification model and the category labeling result corresponding to the historical haze feature map at the historical time are obtained to obtain the corresponding loss function, which specifically includes:

[0025] According to the category labeling results corresponding to the historical haze feature map at the historical time, the classification tasks corresponding to each combination of features are determined;

[0026] Determine a calculation formula for a loss function according to the classification task;

[0027] According to the determined calculation formula of the loss function, the loss function corresponding to each combination feature is obtained respectively.

[0028] The beneficial effects of the invention of this method are as follows: a method for detecting haze distribution and aerosol optical thickness based on satellite remote sensing is proposed, including obtaining satellite remote sensing data of a target monitoring area, dividing the target monitoring area into multiple target areas, performing inversion calculations on the satellite remote sensing data of each target area, and determining the aerosol optical thickness of each target area; obtaining the hyperspectral image data of each target area according to the satellite remote sensing data of each target area, and obtaining the haze characteristic map of each target area according to the hyperspectral image data of each target area, and performing spatial matching between the aerosol optical thickness of each target area and the haze characteristic map of the target area; using an artificial neural network to classify the aerosol optical thickness of each target area and the haze characteristic map of the target area, and obtaining the haze distribution information of the target monitoring area. The satellite remote sensing of the present invention can provide haze distribution and aerosol optical thickness detection in a wide range, and has a wide range of applications in pollutant monitoring, determination of pollution events, pollution source analysis, and regional transportation of pollutants.

[0029] Another technical solution of the present invention to solve the above technical problem is as follows:

[0030] A haze distribution and aerosol optical thickness detection device based on satellite remote sensing, comprising:

[0031] An acquisition module is used to obtain satellite remote sensing data of a target monitoring area, divide the target monitoring area into multiple target areas, perform inversion calculations on the satellite remote sensing data of each target area, and determine the aerosol optical thickness of each target area;

[0032] A feature module is used to obtain the hyperspectral image data of each target area according to the satellite remote sensing data of each target area, and obtain the haze characteristic map of each target area according to the hyperspectral image data of each target area, and spatially match the aerosol optical thickness of each target area with the haze characteristic map of the target area;

[0033] The classification module is used to classify the aerosol optical thickness of each target area and the haze characteristic map of the target area by using an artificial neural network to obtain the haze distribution information of the target monitoring area.

[0034] Furthermore, the feature module is specifically used to perform normalization processing after smoothing and filtering the hyperspectral image data and removing spectral line noise to obtain pre-processed hyperspectral image data;

[0035] The pre-processed hyperspectral image data is subjected to a weighted grayscale algorithm to obtain a grayscale value of the hyperspectral image data;

[0036] A haze characteristic map of the target area is obtained according to the preprocessed hyperspectral image data and the grayscale value of the hyperspectral image.

[0037] Furthermore, the classification module is specifically used to extract features of a preset type from the haze feature map, and construct a combined feature corresponding to the haze feature map based on the extracted features;

[0038] The combined features corresponding to the haze feature map and the aerosol optical thickness of the target area are used as inputs of an artificial neural network model, and the haze distribution information of the target monitoring area is determined according to the output result of the artificial neural network model.

[0039] Furthermore, it also includes a model building module for using the category labeling results corresponding to the historical haze characteristic graphs at each historical time as training data;

[0040] Extracting a preset number of preset type features from the historical haze feature map, and obtaining a combination feature corresponding to the historical haze feature map based on the extracted type features;

[0041] Taking the combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map as inputs of a classification model, and obtaining output results of the classification model for the combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map;

[0042] Obtaining a corresponding loss function according to the output result of the classification model for the aerosol optical thickness corresponding to the historical time of the historical haze characteristic map and the historical time, and the category labeling result corresponding to the historical haze characteristic map at the historical time;

[0043] According to the loss function, the loss function of the classification model is determined, and the loss function of the classification model is used to adjust the parameters of the classification model to obtain the artificial neural network model.

[0044] Furthermore, the model building module is specifically used to determine the classification task corresponding to each combination feature according to the category labeling result corresponding to the historical haze feature map at the historical time;

[0045] Determine a calculation formula for a loss function according to the classification task;

[0046] According to the determined calculation formula of the loss function, the loss function corresponding to each combination feature is obtained respectively.

[0047] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments of the present invention or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 A schematic diagram of a flow chart of a method for detecting haze distribution and aerosol optical thickness based on satellite remote sensing according to an embodiment of the present invention;

[0050] Figure 2 This is a module schematic diagram of a device for detecting haze distribution and aerosol optical thickness based on satellite remote sensing according to another embodiment of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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 of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0052] like Figure 1 As shown, a method for detecting haze distribution and aerosol optical thickness based on satellite remote sensing according to an embodiment of the present invention comprises the following steps:

[0053] 110. Obtain satellite remote sensing data of a target monitoring area, divide the target monitoring area into a plurality of target areas, perform inversion calculations on the satellite remote sensing data of each target area, and determine the aerosol optical thickness of each target area.

[0054] 120. Obtain hyperspectral image data of each target area based on the satellite remote sensing data of each target area, and obtain haze characteristic maps of each target area based on the hyperspectral image data of each target area, and spatially match the aerosol optical thickness of each target area with the haze characteristic maps of the target areas.

[0055] 130. Using an artificial neural network, classify the aerosol optical thickness of each target area and the haze characteristic map of the target area to obtain haze distribution information of the target monitoring area.

[0056] Furthermore, in step 120, the haze characteristic map of each target area is obtained according to the hyperspectral image data of each target area, which specifically includes:

[0057] The hyperspectral image data is smoothed and filtered, spectral line noise is removed, and then normalized to obtain pre-processed hyperspectral image data.

[0058] The preprocessed hyperspectral image data is subjected to a weighted grayscale algorithm to obtain a grayscale value of the hyperspectral image data.

[0059] A haze characteristic map of the target area is obtained according to the preprocessed hyperspectral image data and the grayscale value of the hyperspectral image.

[0060] Furthermore, in step 130, the aerosol optical thickness of each target area and the haze characteristic map of the target area are classified by using an artificial neural network to obtain the haze distribution information of the target monitoring area, which specifically includes:

[0061] Features of a preset type are extracted from the haze feature map, and combined features corresponding to the haze feature map are constructed based on the extracted features.

[0062] The combined features corresponding to the haze feature map and the aerosol optical thickness of the target area are used as inputs of an artificial neural network model, and the haze distribution information of the target monitoring area is determined according to the output result of the artificial neural network model.

[0063] Furthermore, the artificial neural network model is established by:

[0064] The category labeling results corresponding to the historical haze characteristic maps at each historical time are used as training data;

[0065] A preset number of preset type features are extracted from the historical haze feature map, and based on the extracted type features, a combination feature corresponding to the historical haze feature map is obtained.

[0066] The combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map are used as inputs of a classification model, and an output result of the classification model for the combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map is obtained.

[0067] A corresponding loss function is obtained according to the output result of the classification model of the aerosol optical thickness corresponding to the historical time of the historical haze characteristic map and the historical time, and the category labeling result corresponding to the historical haze characteristic map at the historical time.

[0068] According to the loss function, the loss function of the classification model is determined, and the loss function of the classification model is used to adjust the parameters of the classification model to obtain the artificial neural network model.

[0069] Furthermore, according to the output result of the classification model for the aerosol optical thickness corresponding to the historical time of the combined feature and the historical haze feature map and the category labeling result corresponding to the historical haze feature map at the historical time, a corresponding loss function is obtained, which specifically includes:

[0070] According to the category labeling results corresponding to the historical haze feature maps at historical times, the classification tasks corresponding to each combination of features are determined.

[0071] A calculation formula for the loss function is determined according to the classification task.

[0072] According to the determined calculation formula of the loss function, the loss function corresponding to each combination feature is obtained respectively.

[0073] A method for detecting haze distribution and aerosol optical thickness based on satellite remote sensing proposed based on the above-mentioned embodiment includes obtaining satellite remote sensing data of a target monitoring area, dividing the target monitoring area into multiple target areas, performing inversion calculations on the satellite remote sensing data of each target area, and determining the aerosol optical thickness of each target area; obtaining hyperspectral image data of each target area according to the satellite remote sensing data of each target area, and obtaining haze characteristic maps of each target area according to the hyperspectral image data of each target area, and spatially matching the aerosol optical thickness of each target area with the haze characteristic map of the target area; using an artificial neural network to classify the aerosol optical thickness of each target area and the haze characteristic map of the target area, and obtaining haze distribution information of the target monitoring area. The satellite remote sensing of the present invention can provide haze distribution and aerosol optical thickness detection in a wide range, and has a wide range of applications in pollutant monitoring, determination of pollution events, pollution source analysis, and regional transportation of pollutants.

[0074] like Figure 2As shown, a haze distribution and aerosol optical thickness detection device based on satellite remote sensing includes:

[0075] The acquisition module is used to obtain satellite remote sensing data of the target monitoring area, divide the target monitoring area into multiple target areas, perform inversion calculations on the satellite remote sensing data of each target area, and determine the aerosol optical thickness of each target area.

[0076] The feature module is used to obtain the hyperspectral image data of each target area according to the satellite remote sensing data of each target area, and obtain the haze characteristic map of each target area according to the hyperspectral image data of each target area, and spatially match the aerosol optical thickness of each target area with the haze characteristic map of the target area.

[0077] The classification module is used to classify the aerosol optical thickness of each target area and the haze characteristic map of the target area by using an artificial neural network to obtain the haze distribution information of the target monitoring area.

[0078] Furthermore, the feature module is specifically used to perform normalization processing after smoothing and filtering the hyperspectral image data and removing spectral line noise to obtain pre-processed hyperspectral image data;

[0079] The pre-processed hyperspectral image data is subjected to a weighted grayscale algorithm to obtain a grayscale value of the hyperspectral image data;

[0080] A haze characteristic map of the target area is obtained according to the preprocessed hyperspectral image data and the grayscale value of the hyperspectral image.

[0081] Furthermore, the classification module is specifically used to extract features of a preset type from the haze feature map, and construct a combined feature corresponding to the haze feature map based on the extracted features;

[0082] The combined features corresponding to the haze feature map and the aerosol optical thickness of the target area are used as inputs of an artificial neural network model, and the haze distribution information of the target monitoring area is determined according to the output result of the artificial neural network model.

[0083] Furthermore, it also includes a model building module for using the category labeling results corresponding to the historical haze characteristic graphs at each historical time as training data;

[0084] Extracting a preset number of preset type features from the historical haze feature map, and obtaining a combination feature corresponding to the historical haze feature map based on the extracted type features;

[0085] Taking the combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map as inputs of a classification model, and obtaining output results of the classification model for the combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map;

[0086] Obtaining a corresponding loss function according to the output result of the classification model for the aerosol optical thickness corresponding to the historical time of the historical haze characteristic map and the historical time, and the category labeling result corresponding to the historical haze characteristic map at the historical time;

[0087] According to the loss function, the loss function of the classification model is determined, and the loss function of the classification model is used to adjust the parameters of the classification model to obtain the artificial neural network model.

[0088] Furthermore, the model building module is specifically used to determine the classification task corresponding to each combination feature according to the category labeling result corresponding to the historical haze feature map at the historical time;

[0089] Determine a calculation formula for a loss function according to the classification task;

[0090] According to the determined calculation formula of the loss function, the loss function corresponding to each combination feature is obtained respectively.

[0091] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0092] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0093] In the embodiments provided by the present invention, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0094] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0095] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0096] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0097] Based on such understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of each of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0098] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

[0099] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for detecting haze distribution and aerosol optical thickness based on satellite remote sensing, characterized in that: include: Acquire satellite remote sensing data of a target monitoring area, divide the target monitoring area into a plurality of target areas, perform inversion calculations on the satellite remote sensing data of each target area, and determine the aerosol optical thickness of each target area; According to the satellite remote sensing data of each target area, the hyperspectral image data of each target area is obtained, and according to the hyperspectral image data of each target area, the haze characteristic map of each target area is obtained, and the aerosol optical thickness of each target area and the haze characteristic map of the target area are spatially matched; Using an artificial neural network to classify the aerosol optical thickness of each target area and the haze characteristic map of the target area, so as to obtain the haze distribution information of the target monitoring area; The step of obtaining the haze characteristic map of each target area according to the hyperspectral image data of each target area specifically includes: The hyperspectral image data is smoothed and filtered, spectral line noise is removed, and then normalized to obtain pre-processed hyperspectral image data; The pre-processed hyperspectral image data is subjected to a weighted grayscale algorithm to obtain a grayscale value of the hyperspectral image data; Obtaining a haze characteristic map of the target area according to the preprocessed hyperspectral image data and the grayscale value of the hyperspectral image; The method of using an artificial neural network to classify the aerosol optical thickness of each target area and the haze characteristic map of the target area to obtain the haze distribution information of the target monitoring area specifically includes: Extracting features of a preset type from the haze feature map, and constructing combined features corresponding to the haze feature map based on the extracted features; Using the combined features corresponding to the haze feature map and the aerosol optical thickness of the target area as inputs of an artificial neural network model, and determining haze distribution information of the target monitoring area according to an output result of the artificial neural network model; The artificial neural network model is established by: The category labeling results corresponding to the historical haze characteristic maps at each historical time are used as training data; Extracting a preset number of preset type features from the historical haze feature map, and obtaining a combination feature corresponding to the historical haze feature map based on the extracted type features; Taking the combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map as inputs of a classification model, and obtaining output results of the classification model for the combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map; Obtaining a corresponding loss function according to the output result of the classification model for the aerosol optical thickness corresponding to the historical time of the historical haze characteristic map and the historical time, and the category labeling result corresponding to the historical haze characteristic map at the historical time; According to the loss function, the loss function of the classification model is determined, and the loss function of the classification model is used to adjust the parameters of the classification model to obtain the artificial neural network model.

2. The method for detecting haze distribution and aerosol optical thickness based on satellite remote sensing according to claim 1, characterized in that: The output result of the aerosol optical thickness corresponding to the historical time of the historical haze feature map and the historical time of the classification model is obtained according to the category labeling result corresponding to the historical haze feature map at the historical time, and the corresponding loss function is obtained, specifically including: According to the category labeling results corresponding to the historical haze feature map at the historical time, the classification tasks corresponding to each combination of features are determined; Determine a calculation formula for a loss function according to the classification task; According to the determined calculation formula of the loss function, the loss function corresponding to each combination feature is obtained respectively.

3. A haze distribution and aerosol optical thickness detection device based on satellite remote sensing, characterized in that: include: An acquisition module is used to obtain satellite remote sensing data of a target monitoring area, divide the target monitoring area into multiple target areas, perform inversion calculations on the satellite remote sensing data of each target area, and determine the aerosol optical thickness of each target area; A feature module is used to obtain the hyperspectral image data of each target area according to the satellite remote sensing data of each target area, and obtain the haze characteristic map of each target area according to the hyperspectral image data of each target area, and spatially match the aerosol optical thickness of each target area with the haze characteristic map of the target area; A classification module, used to classify the aerosol optical thickness of each target area and the haze characteristic map of the target area by using an artificial neural network to obtain haze distribution information of the target monitoring area; The feature module is specifically used to perform smoothing filtering and spectral line noise removal on the hyperspectral image data and then perform normalization processing to obtain pre-processed hyperspectral image data; The pre-processed hyperspectral image data is subjected to a weighted grayscale algorithm to obtain a grayscale value of the hyperspectral image data; Obtaining a haze characteristic map of the target area according to the preprocessed hyperspectral image data and the grayscale value of the hyperspectral image; The classification module is specifically used to extract features of a preset type from the haze feature map, and construct a combined feature corresponding to the haze feature map based on the extracted features; Using the combined features corresponding to the haze feature map and the aerosol optical thickness of the target area as inputs of an artificial neural network model, and determining haze distribution information of the target monitoring area according to an output result of the artificial neural network model; It also includes a model building module for using the category labeling results corresponding to the historical haze characteristic maps at each historical time as training data; Extracting a preset number of preset type features from the historical haze feature map, and obtaining a combination feature corresponding to the historical haze feature map based on the extracted type features; Taking the combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map as inputs of a classification model, and obtaining output results of the classification model for the combined feature and the aerosol optical thickness corresponding to the historical time of the historical haze feature map; Obtaining a corresponding loss function according to the output result of the classification model for the aerosol optical thickness corresponding to the historical time of the historical haze characteristic map and the historical time, and the category labeling result corresponding to the historical haze characteristic map at the historical time; Determining the loss function of the classification model according to the loss function, and adjusting the parameters of the classification model using the loss function of the classification model to obtain the artificial neural network model; The model building module is specifically used to determine the classification tasks corresponding to each combination feature according to the category labeling results corresponding to the historical haze feature map at the historical time; Determine a calculation formula for a loss function according to the classification task; According to the determined calculation formula of the loss function, the loss function corresponding to each combination feature is obtained respectively.

Citation Information

Patent Citations

  • Atmospheric pollution satellite remote sensing monitoring method and device and computer readable medium

    CN109001091A

  • Arid region dust-haze pollution early warning method based on remote sensing technology

    CN109406361A

  • High-resolution remote sensing image weak target detection method based on deep learning

    CN110728658A

  • Offshore haze remote sensing identification method and system under clear sky condition

    CN111539366A

  • Method and device for detecting cloud cover in hyperspectral satellite image

    CN113096129A