Method and device for discovering pollution discharge case source based on remote sensing image

Abstract

The invention discloses a pollution discharge case source discovery method and device based on a remote sensing image, and belongs to the technical field of satellite remote sensing. The method comprises the following steps: acquiring a to-be-processed remote sensing image covering a to-be-studied area and a background remote sensing image, and performing preprocessing; performing water quality parameter inversion by using the water quality parameter remote sensing inversion model to obtain a water quality parameter inversion result; obtaining a water quality parameter high value area according to the water quality parameter inversion result and the water quality parameter threshold value; subtracting the water quality parameter high-value region of the background remote sensing image from the water quality parameter high-value region of the remote sensing image to be processed to obtain a water quality change abnormal region; and performing pollution traceability analysis according to the water quality change abnormal region to obtain a pollution discharge source list. According to the invention, accurate monitoring of water quality change abnormal areas is realized, and pollution discharge sources can be efficiently and accurately found in real time.

Description

Technical Field

[0001] The present invention relates to the field of satellite remote sensing technology, and in particular to a method and device for discovering pollution sources based on remote sensing images. Background Art

[0002] As an integral part of the Earth's water cycle, the water quality and ecological status of rivers and lakes are directly related to human production and life, as well as the health and stability of the natural environment. With global climate change and the intensification of human activities, river and lake regions face a series of serious problems, such as water pollution and ecological degradation. Monitoring the water quality, ecological status, and changing trends of river and lake regions has become a key task in environmental protection and water resources management. In response to the damage to water resources and the ecological environment caused by sewage discharge, traditional river and lake water quality monitoring methods, such as on-site sampling and laboratory analysis, can provide relatively accurate water quality and ecological data. However, they are time-consuming and costly. Limited by time, manpower, and material resources, they are difficult to achieve large-scale, high-frequency real-time monitoring. Furthermore, long-term retrospective investigations and evidence collection are difficult.

[0003] In recent years, the rapid development of remote sensing technology has provided new solutions for river and lake monitoring. Remote sensing images have been widely used in the field of environmental monitoring due to their wide coverage, fast acquisition speed, and large amount of information.

[0004] Existing techniques typically invert water quality parameters using remote sensing imagery. These inverted water quality parameters are then used to identify areas of water quality anomalies, which are then used to analyze possible pollution sources. However, the distribution of pollution sources determined by this method is inaccurate. One reason is that the inverted water quality parameters are influenced not only by the pollution source but also by the water quality of the water body itself. This results in inaccurate extraction of anomaly areas, which in turn affects the accuracy of pollution source identification.

[0005] Furthermore, in practical applications, relying on remote sensing imagery for river and lake monitoring still presents challenges. The water quality and ecological status of river and lake regions are affected by a variety of factors, such as human activities, and changes in these factors can lead to significant changes in the water quality and ecological status of river and lake regions.

[0006] Therefore, in practical applications, how to effectively use remote sensing image data to achieve precise monitoring of high-value water quality areas in rivers and lakes and to discover pollution sources efficiently, accurately and in real time is still an urgent problem to be solved. Summary of the Invention

[0007] In order to solve the above technical problems, the present invention provides a method and device for discovering pollution sources based on remote sensing images, which realizes precise monitoring of areas with abnormal water quality changes and can discover pollution sources efficiently, accurately and in real time.

[0008] The present invention provides the following technical solutions:

[0009] A method for discovering pollution sources based on remote sensing images, comprising:

[0010] S1: Acquire the remote sensing images to be processed and the background remote sensing images covering the area to be studied;

[0011] The background remote sensing image is a remote sensing image of the area to be studied during a period of no human activity or a period of little human activity;

[0012] S2: Preprocessing the remote sensing image to be processed and the background remote sensing image;

[0013] S3: using the constructed water quality parameter remote sensing inversion model, performing water quality parameter inversion on the remote sensing image to be processed and the background remote sensing image, and obtaining water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image;

[0014] S4: obtaining high-value areas of water quality parameters of the remote sensing image to be processed and the background remote sensing image according to the water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image and a preset water quality parameter threshold;

[0015] S5: subtracting the high-value area of ​​the water quality parameter of the background remote sensing image from the high-value area of ​​the water quality parameter of the remote sensing image to be processed to obtain the area with abnormal water quality change;

[0016] S6: Conduct pollution source analysis based on the areas with abnormal water quality changes to obtain a list of pollution sources.

[0017] Furthermore, the remote sensing images to be processed and the background remote sensing images are multi-source remote sensing images, and the multi-source remote sensing images include multi-temporal, multi-spectral, multi-sensor, multi-platform and / or multi-resolution remote sensing image data.

[0018] Furthermore, the preprocessing includes atmospheric correction, geometric correction and radiation correction.

[0019] Furthermore, a water quality parameter remote sensing inversion model is constructed by the following method:

[0020] Based on the measured water quality parameters and remote sensing image reflectance data, the water quality parameter remote sensing inversion model is constructed using a semi-empirical model or a machine learning algorithm.

[0021] Furthermore, the water quality parameter is the total phosphorus concentration, and a water quality parameter remote sensing inversion model is constructed by the following method:

[0022] The total phosphorus concentration was inverted using the ratio of the remote sensing image bands and the XGBoost model, or multiple linear regression was performed to obtain the remote sensing inversion model of water quality parameters corresponding to the total phosphorus concentration.

[0023] Furthermore, the water quality parameter is suspended matter concentration, and the water quality parameter remote sensing inversion model performs suspended matter concentration inversion using the following formula:

[0024]

[0025] Where T is the suspended solids concentration, R red 、R Green and R Nir They are the reflectance data of the red band, green band and near-infrared band of the remote sensing image to be processed or the background remote sensing image.

[0026] Furthermore, the S6 includes:

[0027] S61: Extracting the distribution and area of ​​the abnormal water quality change area;

[0028] S62: Acquire land use data in and around the area where the water quality changes abnormally;

[0029] S63: Using the distribution and area of ​​the areas with abnormal water quality changes and the land use data, conduct pollution source tracing analysis to obtain a list of pollution sources.

[0030] A pollution source discovery device based on remote sensing images, comprising:

[0031] A data acquisition module is used to obtain remote sensing images to be processed and background remote sensing images covering the area to be studied;

[0032] The background remote sensing image is a remote sensing image of the area to be studied during a period of no human activity or a period of little human activity;

[0033] A preprocessing module, used for preprocessing the remote sensing image to be processed and the background remote sensing image;

[0034] A water quality inversion module is used to perform water quality parameter inversion on the remote sensing image to be processed and the background remote sensing image using the constructed water quality parameter remote sensing inversion model, and obtain water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image;

[0035] A high-value area determination module is used to obtain high-value areas of water quality parameters of the remote sensing image to be processed and the background remote sensing image based on the water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image and a preset water quality parameter threshold;

[0036] An abnormal area determination module is used to subtract the high-value area of ​​the water quality parameter of the background remote sensing image from the high-value area of ​​the water quality parameter of the remote sensing image to be processed to obtain the area with abnormal water quality changes;

[0037] The pollution source tracing module is used to conduct pollution source tracing analysis based on the abnormal water quality change area and obtain a list of pollution source cases.

[0038] Furthermore, the remote sensing images to be processed and the background remote sensing images are multi-source remote sensing images, and the multi-source remote sensing images include multi-temporal, multi-spectral, multi-sensor, multi-platform and / or multi-resolution remote sensing image data.

[0039] Furthermore, the preprocessing includes atmospheric correction, geometric correction and radiation correction.

[0040] Furthermore, a water quality parameter remote sensing inversion model is constructed through the following process:

[0041] Based on the measured water quality parameters and remote sensing image reflectance data, the water quality parameter remote sensing inversion model is constructed using a semi-empirical model or a machine learning algorithm.

[0042] Furthermore, the water quality parameter is the total phosphorus concentration, and a water quality parameter remote sensing inversion model is constructed through the following process:

[0043] The total phosphorus concentration was inverted using the ratio of the remote sensing image bands and the XGBoost model, or multiple linear regression was performed to obtain the remote sensing inversion model of water quality parameters corresponding to the total phosphorus concentration.

[0044] Furthermore, the water quality parameter is suspended matter concentration, and the water quality parameter remote sensing inversion model performs suspended matter concentration inversion using the following formula:

[0045]

[0046] Where T is the suspended solids concentration, R red 、R Green and R Nir They are the reflectance data of the red band, green band and near-infrared band of the remote sensing image to be processed or the background remote sensing image.

[0047] Furthermore, the pollution tracing module includes:

[0048] An abnormal area extraction unit, used to extract the distribution and area of ​​the abnormal water quality change area;

[0049] A land use data acquisition unit, configured to acquire land use data in and around the area where the water quality changes abnormally;

[0050] The pollution source tracing unit is used to use the distribution and area of ​​abnormal water quality change areas and the land use data to conduct pollution source tracing analysis and obtain a list of pollution source cases.

[0051] The present invention has the following beneficial effects:

[0052] This invention utilizes remote sensing technology to achieve real-time monitoring of pollution sources in water bodies, providing timely information on the water quality and ecological status of these areas. This improves monitoring efficiency, reduces monitoring costs, and increases monitoring effectiveness compared to traditional monitoring methods. This invention uses remote sensing imagery data from periods of low or no human activity as a comparison baseline. During these periods, water quality is less affected by factors such as human activity, and can better capture changes in the water quality of the water body itself. Comparative analysis with currently processed remote sensing imagery data can more accurately reveal changing trends in water quality under the influence of factors such as human activity, particularly changes in high-value areas. This comparative analysis method minimizes the impact of the water quality of rivers and lakes, enabling precise source tracing and analysis. This supports pollution source identification and improves the accuracy of pollution source lists, thereby enhancing the precision and reliability of water quality monitoring in rivers and lakes. This invention provides a scientific basis and decision-making support for water management and protection, helping to identify potential environmental issues and facilitating the development of measures tailored to the water quality of rivers and lakes and the impact of human activity, thereby improving water quality and the ecological environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is a flow chart of the method for discovering pollution sources based on remote sensing images of the present invention;

[0054] Figure 2 This is a schematic diagram of the pollution source discovery device based on remote sensing images of the present invention. DETAILED DESCRIPTION

[0055] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0056] Example 1:

[0057] The embodiment of the present invention provides a method for discovering pollution sources based on remote sensing images, such as Figure 1 As shown, the method includes:

[0058] S1: Obtain the remote sensing image to be processed and the background remote sensing image covering the area to be studied.

[0059] Among them, the background remote sensing image is the remote sensing image of the area to be studied during the period of no human activity or little human activity.

[0060] The study area in the present invention generally refers to an area containing water bodies such as lakes, rivers, and reservoirs. The remote sensing images during periods of no or minimal human activity described in the embodiments of the present invention refer to remote sensing images taken during a period when the study area was unaffected or minimally affected by human activity, serving as background remote sensing images. For example, remote sensing images from 2020 to 2022 are used. Background remote sensing images eliminate the effects of human activity on water quality and reflect the water quality of the water itself.

[0061] In practice, the acquisition and processing of remote sensing image data presents certain technical difficulties and cost constraints. To address these issues, the present invention uses multi-source remote sensing images, including those from multiple temporal, multi-spectral, multi-sensor, multi-platform (e.g., satellite, drone, or other remote sensing platforms), and / or multi-resolution remote sensing image data, for both the processed and background remote sensing images.

[0062] Multi-source remote sensing image data encompasses a wealth of information spanning multiple temporal, multi-spectral, multi-sensor, multi-platform, and multi-resolution levels (e.g., meter-level, sub-meter-level, etc.). Remote sensing image data can be obtained from multiple sources, addressing the technical difficulties and cost constraints of acquiring and processing remote sensing image data. Furthermore, multi-source remote sensing image data can fuse image information at different temporal, spatial, and spectral resolutions, providing a richer and more comprehensive data source for river and lake monitoring. Effectively integrating and utilizing this multi-source data is crucial for a comprehensive and in-depth understanding of water quality, ecological conditions, and changing trends in river and lake regions. This facilitates the accurate identification and real-time monitoring of pollution discharges, as well as the tracing and precise investigation of pollution sources.

[0063] S2: Preprocess the remote sensing images to be processed and the background remote sensing images.

[0064] Preprocessing includes atmospheric correction, geometric correction, radiometric correction, and image fusion (for multi-source remote sensing images). The goal of preprocessing is to improve the quality and usability of remote sensing data. Both the processed and baseline remote sensing images undergo the same preprocessing to ensure comparability.

[0065] S3: Using the constructed water quality parameter remote sensing inversion model, water quality parameters are inverted for the remote sensing image to be processed and the background remote sensing image to obtain the water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image.

[0066] In the present invention, a water quality parameter remote sensing inversion model can be constructed by the following method:

[0067] Based on measured water quality parameters (such as suspended matter concentration, total phosphorus, etc.) and remote sensing image reflectance data, a water quality parameter remote sensing inversion model is constructed using semi-empirical models or machine learning algorithms.

[0068] For example, when the water quality parameter is total phosphorus concentration, the water quality parameter remote sensing inversion model can be constructed by the following method:

[0069] The total phosphorus concentration was inverted using the ratio of the remote sensing image bands and the XGBoost model, or multiple linear regression was performed to obtain the remote sensing inversion model of water quality parameters corresponding to the total phosphorus concentration.

[0070] After obtaining the water quality parameter remote sensing inversion model, water quality parameter inversion is performed to obtain water quality parameter inversion results for the processed remote sensing image and the background remote sensing image. If there is only one water quality parameter, the inversion results for that single water quality parameter are obtained for the processed remote sensing image and the background remote sensing image respectively. If there are multiple water quality parameters, the inversion results for multiple water quality parameters are obtained for the processed remote sensing image and the background remote sensing image respectively.

[0071] In one example, when the water quality parameter is suspended solids concentration, there is a certain correlation between the suspended solids concentration and the difference between the green light band and the red light band, and the difference between the green light band and the near-infrared band. The present invention uses a multi-band combination method to invert the suspended solids concentration. Specifically, the water quality parameter remote sensing inversion model inverts the suspended solids concentration using the following formula:

[0072]

[0073] Where T is the suspended solids concentration, R red 、R Green and R Nir They are the reflectance data of the red band, green band and near-infrared band of the remote sensing image to be processed or the background remote sensing image.

[0074] The water quality parameter inversion results from the processed remote sensing imagery represent the actual water quality parameter values ​​at the time of monitoring, while the water quality parameter inversion results from the background remote sensing imagery represent the water quality parameter values ​​of the water body itself. By comparing and analyzing the water quality parameter inversion results from these two types of remote sensing images, pollution source tracing can be achieved.

[0075] S4: According to the water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image and the pre-set water quality parameter thresholds, the high-value areas of the water quality parameters of the remote sensing image to be processed and the background remote sensing image are obtained.

[0076] This step is used to screen out areas with high water quality parameter values ​​in both the processed and baseline remote sensing images. This screening method uses a threshold comparison method. Relevant thresholds can be set based on relevant standards. For example, according to the "Surface Water Environmental Quality Standard GB 3838-2002," areas with a total phosphorus concentration threshold worse than Class III can be screened out. Alternatively, areas with high suspended solids concentrations can be screened out based on the "Comprehensive Wastewater Discharge Standard GB8978-1996," using the Level 1 standard.

[0077] S5: Subtract the high-value area of ​​water quality parameters in the background remote sensing image from the high-value area of ​​water quality parameters in the remote sensing image to be processed to obtain the area with abnormal water quality changes.

[0078] This step is based on a comparative analysis of the high-value areas of water quality parameters in the background remote sensing images during periods with no or little human activity and the high-value areas of water quality parameters in the current remote sensing images to be processed, and preliminarily identifies areas with abnormal changes in river and lake water quality during periods with no or little human activity, as well as areas with abnormal river and lake water quality during periods of human activity, that is, areas with severe water pollution.

[0079] Specifically, the high-value area of ​​the water quality parameter of the remote sensing image to be processed can be subtracted from the high-value area of ​​the water quality parameter of the background remote sensing image to obtain the area with abnormal water quality changes.

[0080] S6: Conduct pollution source analysis based on areas with abnormal water quality changes and obtain a list of pollution sources.

[0081] In the present invention, pollution source tracing analysis can be performed according to any pollution source tracing analysis method or model disclosed in the prior art. The specific source tracing analysis method is not described in detail in this invention. The obtained pollution source list includes the number and location of the pollution source.

[0082] As an improvement to the embodiment of the present invention, the aforementioned S6 includes:

[0083] S61: Extract the distribution and area of ​​abnormal water quality change areas.

[0084] S62: Obtain land use data in and around the area with abnormal water quality changes.

[0085] Land use data is used to determine the impact of surrounding land use on areas experiencing abnormal water quality changes. This data can be obtained from platforms such as the National Geographic Information Center, or by interpreting it using sub-meter high-resolution remote sensing imagery.

[0086] S63: Use the distribution and area of ​​abnormal water quality change areas and land use data to conduct pollution source analysis and obtain a list of pollution sources.

[0087] The present invention utilizes the distribution and area of ​​regions with abnormal water quality changes, as well as land use data, to conduct pollution source tracing analysis on changes in areas with high water quality parameters. Combined with factors such as the surrounding environment of the water area and human activities, the present invention analyzes possible natural and human factors that may cause water environment damage, focusing on investigating the potential connection between human activities and water environment damage, and generating a list of pollution sources during the period of human activity.

[0088] Furthermore, the present invention can establish a continuous monitoring mechanism to provide long-term, stable monitoring of the duration and development of high-quality water quality zones, tracking their changing trends and influencing factors. Finally, based on the monitoring results, a remote sensing monitoring report on wastewater discharge is generated, providing a scientific basis and decision-making support for river and lake management and protection.

[0089] This invention utilizes remote sensing technology to achieve real-time monitoring of pollution sources in water bodies, providing timely information on the water quality and ecological status of these areas. This improves monitoring efficiency, reduces monitoring costs, and increases monitoring effectiveness compared to traditional monitoring methods. This invention uses remote sensing imagery data from periods of low or no human activity as a comparison baseline. During these periods, water quality is less affected by factors such as human activity, and can better capture changes in the water quality of the water body itself. Comparative analysis with currently processed remote sensing imagery data can more accurately reveal changing trends in water quality under the influence of factors such as human activity, particularly changes in high-value areas. This comparative analysis method minimizes the impact of the water quality of rivers and lakes, enabling precise source tracing and analysis. This supports pollution source identification and improves the accuracy of pollution source lists, thereby enhancing the precision and reliability of water quality monitoring in rivers and lakes. This invention provides a scientific basis and decision-making support for water management and protection, helping to identify potential environmental issues and facilitating the development of measures tailored to the water quality of rivers and lakes and the impact of human activity, thereby improving water quality and the ecological environment.

[0090] Example 2:

[0091] The embodiment of the present invention provides a device for discovering pollution sources based on remote sensing images, such as Figure 2 As shown, the device includes:

[0092] The data acquisition module 1 is used to acquire the remote sensing images to be processed and the background remote sensing images covering the area to be studied.

[0093] Among them, the background remote sensing image is the remote sensing image of the area to be studied during the period of no human activity or little human activity.

[0094] The preprocessing module 2 is used to preprocess the remote sensing images to be processed and the background remote sensing images.

[0095] The water quality inversion module 3 is used to use the constructed water quality parameter remote sensing inversion model to invert the water quality parameters of the remote sensing image to be processed and the background remote sensing image, and obtain the water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image.

[0096] The high-value area determination module 4 is used to obtain the high-value areas of water quality parameters of the remote sensing image to be processed and the background remote sensing image based on the water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image and the preset water quality parameter thresholds.

[0097] The abnormal area determination module 5 is used to subtract the high-value area of ​​water quality parameters in the background remote sensing image from the high-value area of ​​water quality parameters in the remote sensing image to be processed to obtain the area with abnormal water quality changes.

[0098] The pollution source tracing module 6 is used to conduct pollution source tracing analysis based on the areas with abnormal water quality changes and obtain a list of pollution sources.

[0099] In one example, the aforementioned remote sensing images to be processed and background remote sensing images are multi-source remote sensing images, which include multi-temporal, multi-spectral, multi-sensor, multi-platform and / or multi-resolution remote sensing image data.

[0100] In the present invention, the preprocessing includes operations such as atmospheric correction, geometric correction and radiation correction.

[0101] As the present invention, a water quality parameter remote sensing inversion model can be constructed through the following process:

[0102] Based on the measured water quality parameters and remote sensing image reflectance data, a water quality parameter remote sensing inversion model is constructed using a semi-empirical model or machine learning algorithm.

[0103] Specifically, when the water quality parameter is total phosphorus concentration, the water quality parameter remote sensing inversion model can be constructed through the following process:

[0104] The total phosphorus concentration was inverted using the ratio of the remote sensing image bands and the XGBoost model, or multiple linear regression was performed to obtain the remote sensing inversion model of water quality parameters corresponding to the total phosphorus concentration.

[0105] When the water quality parameter is suspended matter concentration, the water quality parameter remote sensing inversion model uses the following formula to invert the suspended matter concentration:

[0106]

[0107] Where T is the suspended solids concentration, R red 、R Green and R Nir They are the reflectance data of the red band, green band and near-infrared band of the remote sensing image to be processed or the background remote sensing image.

[0108] As an improvement to the embodiment of the present invention, the pollution tracing module includes:

[0109] The abnormal area extraction unit is used to extract the distribution and area of ​​abnormal water quality change areas.

[0110] The land use data acquisition unit is used to acquire land use data in and around the area where water quality changes are abnormal.

[0111] The pollution source tracing unit is used to conduct pollution source tracing analysis using the distribution and area of ​​areas with abnormal water quality changes, as well as land use data, to obtain a list of pollution sources.

[0112] The device provided in the embodiment of the present invention has the same implementation principles and technical effects as the aforementioned method embodiment. For the sake of brevity, any matters not mentioned in the device embodiment can be referred to the corresponding contents in the aforementioned method embodiment 1. Those skilled in the art will clearly understand that for the convenience and brevity of description, the specific working processes of the aforementioned devices and units can all refer to the corresponding processes in the aforementioned method embodiment and will not be repeated here.

[0113] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present invention, which are used to illustrate the technical solutions of the present invention rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the above-mentioned embodiments, ordinary technicians in this field should understand that any technician familiar with this technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed by the present invention, or make equivalent replacements for some of the technical features therein; and these modifications, changes 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 covered by the protection scope of the present invention.

Claims

1. A method for discovering pollution sources based on remote sensing images, characterized in that: The method comprises: S1: Acquire the remote sensing images to be processed and the background remote sensing images covering the area to be studied; The background remote sensing image is a remote sensing image of the area to be studied during a period of no human activity or a period of little human activity; S2: Preprocessing the remote sensing image to be processed and the background remote sensing image; S3: using the constructed water quality parameter remote sensing inversion model, performing water quality parameter inversion on the remote sensing image to be processed and the background remote sensing image, and obtaining water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image; S4: obtaining high-value areas of water quality parameters of the remote sensing image to be processed and the background remote sensing image according to the water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image and a preset water quality parameter threshold; S5: subtracting the high-value area of ​​the water quality parameter of the background remote sensing image from the high-value area of ​​the water quality parameter of the remote sensing image to be processed to obtain the area with abnormal water quality change; S6: Conduct pollution source analysis based on the areas with abnormal water quality changes to obtain a list of pollution sources.

2. The method for discovering pollution sources based on remote sensing images according to claim 1 is characterized in that: The remote sensing images to be processed and the background remote sensing images are multi-source remote sensing images, and the multi-source remote sensing images include multi-temporal, multi-spectral, multi-sensor, multi-platform and / or multi-resolution remote sensing image data.

3. The method for discovering pollution sources based on remote sensing images according to claim 1, characterized in that: The preprocessing includes atmospheric correction, geometric correction and radiation correction.

4. The method for discovering pollution sources based on remote sensing images according to claim 1, characterized in that: The water quality parameter remote sensing inversion model is constructed by the following method: Based on the measured water quality parameters and remote sensing image reflectance data, the water quality parameter remote sensing inversion model is constructed using a semi-empirical model or a machine learning algorithm.

5. The method for discovering pollution sources based on remote sensing images according to claim 4 is characterized in that: The water quality parameter is the total phosphorus concentration, and a water quality parameter remote sensing inversion model is constructed by the following method: The total phosphorus concentration was inverted using the ratio of the remote sensing image bands and the XGBoost model, or multiple linear regression was performed to obtain the remote sensing inversion model of water quality parameters corresponding to the total phosphorus concentration.

6. The method for discovering pollution sources based on remote sensing images according to claim 4 is characterized in that: The water quality parameter is the suspended matter concentration. The water quality parameter remote sensing inversion model performs suspended matter concentration inversion using the following formula: Where T is the suspended solids concentration, R red 、R Green and R Nir They are the reflectance data of the red band, green band and near-infrared band of the remote sensing image to be processed or the background remote sensing image.

7. The method for discovering pollution sources based on remote sensing images according to any one of claims 1 to 6, characterized in that: The S6 includes: S61: Extracting the distribution and area of ​​the abnormal water quality change area; S62: Acquire land use data in and around the area where the water quality changes abnormally; S63: Using the distribution and area of ​​the areas with abnormal water quality changes and the land use data, conduct pollution source tracing analysis to obtain a list of pollution sources.

8. A pollution source discovery device based on remote sensing images, characterized in that: The device comprises: A data acquisition module is used to obtain remote sensing images to be processed and background remote sensing images covering the area to be studied; The background remote sensing image is a remote sensing image of the area to be studied during a period of no human activity or a period of little human activity; A preprocessing module, used for preprocessing the remote sensing image to be processed and the background remote sensing image; A water quality inversion module is used to perform water quality parameter inversion on the remote sensing image to be processed and the background remote sensing image using the constructed water quality parameter remote sensing inversion model, and obtain water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image; A high-value area determination module is used to obtain high-value areas of water quality parameters of the remote sensing image to be processed and the background remote sensing image based on the water quality parameter inversion results of the remote sensing image to be processed and the background remote sensing image and a preset water quality parameter threshold; An abnormal area determination module is used to subtract the high-value area of ​​the water quality parameter of the background remote sensing image from the high-value area of ​​the water quality parameter of the remote sensing image to be processed to obtain the area with abnormal water quality changes; The pollution source tracing module is used to conduct pollution source tracing analysis based on the abnormal water quality change area and obtain a list of pollution source cases.

9. The pollution source discovery device based on remote sensing images according to claim 8 is characterized in that: The pollution tracing module includes: An abnormal area extraction unit, used to extract the distribution and area of ​​the abnormal water quality change area; A land use data acquisition unit, configured to acquire land use data in and around the area where the water quality changes abnormally; The pollution source tracing unit is used to use the distribution and area of ​​abnormal water quality change areas and the land use data to conduct pollution source tracing analysis and obtain a list of pollution source cases.

Images