A river water quality inversion method, device and medium based on remote sensing image
Through the river water quality inversion method based on remote sensing images, a water quality monitoring inversion model is established using remote sensing images of different resolutions, which solves the problem of labor-dependent monitoring of complex river water quality, real-time and efficient water quality monitoring and dynamic results acquisition are achieved.
Patent Information
- Application Number
- CN202410550031.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-06
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-05-06
AI Technical Summary
In the prior art, the water quality monitoring of complex rivers relies on labor, resulting in high labor costs and long monitoring cycles, making it difficult to obtain dynamic water quality monitoring results in real time and efficiently.
The river water quality inversion method based on remote sensing images is adopted. By determining the combination of the to-determined bands of remote sensing images of different resolutions, a water quality monitoring inversion model is established, and a medium resolution remote sensing image has a wide coverage range and a rich high-resolution remote sensing image data source is generated. Water quality inversion data reduces dependence on manual sampling.
Real-time and efficient monitoring of the water quality of complex rivers is achieved, labor costs are reduced, flexible monitoring cycles are used, and water quality can be dynamically inverted.
Smart Images

Figure CN118470556B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of water quality monitoring, and in particular to a river water quality inversion method, equipment and medium based on remote sensing images. Background Art
[0002] Water is the source of life and an important resource for human survival and development. In particular, the water quality of inland water bodies is directly related to the safety of human water use. Inland surface water bodies mainly include rivers, lakes and reservoirs, which have functions such as flood control, irrigation and climate regulation. They are an important part of the terrestrial ecosystem and the main source of water for human production and life.
[0003] Ensuring that inland surface water bodies are not polluted and carrying out pollution control is the basis for sustainable development. Water quality monitoring is an important part of monitoring and controlling inland river water pollution. It is particularly important to quickly and accurately grasp the spatial distribution characteristics and dynamic changes of surface water quality. This can provide a scientific basis for the formulation of water pollution prevention and control policies, and is of great significance for guiding the protection of water environment, regulating urban production and domestic sewage discharge, and ensuring the ecological security of inland waters.
[0004] Due to the diverse terrain, different river widths, and large number of tributaries of inland surface water rivers, water quality monitoring currently usually requires manual sampling at complex sites and water quality processing and analysis in laboratories. Manual water quality monitoring has high labor costs, long water quality monitoring cycles, and is difficult to obtain dynamic water quality monitoring results in real time and efficiently, which is not conducive to pollution control work based on real-time river pollution conditions. Summary of the invention
[0005] The embodiments of the present application provide a river water quality inversion method, equipment and medium based on remote sensing images, which are used to solve the current problem that water quality monitoring of complex rivers is overly dependent on manual labor, which consumes manpower costs, has a long monitoring cycle, and is difficult to obtain dynamic water quality monitoring results in real time and efficiently.
[0006] On the one hand, an embodiment of the present application provides a river water quality inversion method based on remote sensing images, the method comprising:
[0007] Based on the river monitoring information in the water quality monitoring work area, determine each first undetermined band combination generated by a first predetermined number of remote sensing image bands corresponding to the same river monitoring point; wherein the first undetermined band combination corresponds to a first resolution remote sensing image;
[0008] Performing correlation analysis on the first reflectivity value sets of each of the first undetermined band combinations and the respectively corresponding measured water quality data, so as to determine a first selected band combination corresponding to the specified water quality indicator dimension according to the correlation analysis result;
[0009] Based on the first selected band combination and the measured water quality data corresponding to the specified water quality indicator dimension, a first water quality monitoring inversion model is generated to determine a plurality of working area points and their corresponding first inverted water quality data according to the first water quality monitoring inversion model and the first resolution remote sensing image;
[0010] Based on the second undetermined band combination corresponding to the river monitoring information and the first inverted water quality data, a second selected band combination is generated to generate a second water quality monitoring inversion model according to the second selected band combination and the first inverted water quality data; wherein the second undetermined band combination includes a combination of different remote sensing image bands corresponding to the second resolution remote sensing image;
[0011] According to the second resolution remote sensing image and the second water quality monitoring inversion model, the second inverted water quality data corresponding to the water quality monitoring work area is determined, and the second inverted water quality data is added to the water quality monitoring distribution map and displayed on the user terminal.
[0012] In one implementation of the present application, before determining the first undetermined band combinations generated by the first predetermined number of remote sensing image bands corresponding to the same river monitoring point based on the river monitoring information in the water quality monitoring work area, the method further includes:
[0013] Acquire work area collection data from the user terminal; the work area collection data at least includes remote sensing image data, water quality measured data, and river system vector data in the water quality monitoring work area;
[0014] The data collected in the working area are preprocessed to obtain the river monitoring information; wherein the river monitoring information at least includes a first-resolution remote sensing image, a second-resolution remote sensing image, and the measured water quality data of multiple river monitoring points in the water quality monitoring working area; the measured water quality data corresponds to one or more water quality indicator dimensions; the resolution of the first-resolution remote sensing image is smaller than that of the second-resolution remote sensing image.
[0015] In one implementation of the present application, based on the river monitoring information in the water quality monitoring work area, determining each first undetermined band combination generated by a first predetermined number of remote sensing image bands corresponding to the same river monitoring point specifically includes:
[0016] Determine the first-resolution remote sensing image of the same river monitoring point according to the river monitoring information;
[0017] The first predetermined number of remote sensing image bands are screened from the remote sensing image bands corresponding to the first resolution remote sensing image, and four arithmetic operations are randomly performed on the screened remote sensing image bands to form corresponding first pending band combinations, until all the remote sensing image bands corresponding to the first resolution remote sensing image are randomly combined to determine the first pending band combinations.
[0018] In one implementation of the present application, a correlation analysis is performed on the first reflectance value sets of each of the first undetermined band combinations and the corresponding water quality measured data, so as to determine the first selected band combination corresponding to the specified water quality indicator dimension according to the correlation analysis result, specifically including:
[0019] Calculate the corresponding reflectance values of each first combination according to the first reflectance value sets respectively corresponding to the first undetermined waveband combinations;
[0020] Input each of the first combined reflectance values and the water quality measured data of the designated water quality index dimension corresponding thereto into a preset Pearson correlation coefficient calculation formula to perform correlation analysis to obtain each first correlation coefficient;
[0021] The first correlation coefficients are compared to determine that the first undetermined band combination corresponding to the maximum first correlation coefficient is the first selected band combination.
[0022] In one implementation of the present application, the method further includes:
[0023] Based on the user's specified operation on the user terminal, the corresponding specified water quality index dimensions are obtained; the specified water quality index dimensions include at least: permanganate index, water temperature, pH, dissolved oxygen concentration, conductivity, turbidity, ammonia nitrogen index, total phosphorus content, and total nitrogen content.
[0024] In one implementation of the present application, based on the first selected band combination and the measured water quality data corresponding to the specified water quality indicator dimension, a first water quality monitoring inversion model is generated, specifically including:
[0025] Determine each of the first selected band combinations corresponding to a plurality of different river monitoring points and each of the first combination reflectance values corresponding to the combinations;
[0026] Determine each of the water quality measured data corresponding to the specified water quality indicator dimension obtained at the corresponding plurality of different river monitoring points;
[0027] Performing linear regression on each of the first combined reflectivity values and each of the water quality measured data to obtain the first water quality monitoring inversion model;
[0028] Determining a plurality of working area points and their corresponding first inverted water quality data according to the first water quality monitoring inversion model and the first resolution remote sensing image, specifically including:
[0029] Based on a point designation operation performed by a user through a user terminal on the remote sensing image of the first resolution, determining a corresponding number of the working area points and their corresponding first combined reflectivity values;
[0030] The first combined reflectivity values corresponding to each of the working area points are respectively input into the first water quality monitoring inversion model, so that the output values of the first water quality monitoring inversion model are used as the first inverted water quality data of the corresponding each of the working area points.
[0031] In an implementation of the present application, calculating the corresponding reflectance values of each first combination according to the first reflectance value sets respectively corresponding to the first undetermined band combinations specifically includes:
[0032] According to the first reflectivity value set and four arithmetic operations combination corresponding to the first undetermined band combination, each first reflectivity value in the first reflectivity value set is calculated according to the four arithmetic operations combination, so as to use the calculation result as the first combination reflectivity value.
[0033] In one implementation of the present application, based on the second undetermined band combination corresponding to the river monitoring information and the first inverted water quality data, a second selected band combination is generated, so as to generate a second water quality monitoring inversion model according to the second selected band combination and the first inverted water quality data, specifically including:
[0034] Randomly combining different remote sensing image bands corresponding to the remote sensing image with the second resolution in the river monitoring information to obtain the second undetermined band combinations;
[0035] Performing correlation analysis on the second combined reflectivity values of each of the second undetermined band combinations and the first inverted water quality data respectively corresponding thereto, so as to generate the corresponding second selected band combination according to the correlation analysis result;
[0036] A linear regression is performed on the second combined reflectivity values corresponding to each of the second selected band combinations and the corresponding first inverted water quality data to obtain the second water quality monitoring inversion model according to the linear regression equation.
[0037] On the other hand, the embodiment of the present application further provides a river water quality inversion device based on remote sensing images, the device comprising:
[0038] at least one processor; and a memory in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to:
[0039] Based on the river monitoring information in the water quality monitoring work area, determine each first undetermined band combination generated by a first predetermined number of remote sensing image bands corresponding to the same river monitoring point; wherein the first undetermined band combination corresponds to a first resolution remote sensing image;
[0040] Performing correlation analysis on the first reflectivity value sets of each of the first undetermined band combinations and the respectively corresponding measured water quality data, so as to determine a first selected band combination corresponding to the specified water quality indicator dimension according to the correlation analysis result;
[0041] Based on the first selected band combination and the measured water quality data corresponding to the specified water quality indicator dimension, a first water quality monitoring inversion model is generated to determine a plurality of working area points and their corresponding first inverted water quality data according to the first water quality monitoring inversion model and the first resolution remote sensing image;
[0042] Based on the second undetermined band combination corresponding to the river monitoring information and the first inverted water quality data, a second selected band combination is generated to generate a second water quality monitoring inversion model according to the second selected band combination and the first inverted water quality data; wherein the second undetermined band combination includes a combination of different remote sensing image bands corresponding to the second resolution remote sensing image;
[0043] According to the second resolution remote sensing image and the second water quality monitoring inversion model, the second inverted water quality data corresponding to the water quality monitoring work area is determined, and the second inverted water quality data is added to the water quality monitoring distribution map and displayed on the user terminal.
[0044] On the other hand, the embodiment of the present application further provides a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured as follows:
[0045] Based on the river monitoring information in the water quality monitoring work area, determine each first undetermined band combination generated by a first predetermined number of remote sensing image bands corresponding to the same river monitoring point; wherein the first undetermined band combination corresponds to a first resolution remote sensing image;
[0046] Performing correlation analysis on the first reflectivity value sets of each of the first undetermined band combinations and the respectively corresponding measured water quality data, so as to determine a first selected band combination corresponding to the specified water quality indicator dimension according to the correlation analysis result;
[0047] Based on the first selected band combination and the measured water quality data corresponding to the specified water quality indicator dimension, a first water quality monitoring inversion model is generated to determine a plurality of working area points and their corresponding first inverted water quality data according to the first water quality monitoring inversion model and the first resolution remote sensing image;
[0048] Based on the second undetermined band combination corresponding to the river monitoring information and the first inverted water quality data, a second selected band combination is generated to generate a second water quality monitoring inversion model according to the second selected band combination and the first inverted water quality data; wherein the second undetermined band combination includes a combination of different remote sensing image bands corresponding to the second resolution remote sensing image;
[0049] According to the second resolution remote sensing image and the second water quality monitoring inversion model, the second inverted water quality data corresponding to the water quality monitoring work area is determined, and the second inverted water quality data is added to the water quality monitoring distribution map and displayed on the user terminal.
[0050] Compared with the prior art, the present invention has the following significant effects:
[0051] Through the above technical solution, the present application can use two types of resolution remote sensing images in the water quality monitoring work area to respectively establish water quality monitoring inversion models. On the one hand, the medium-resolution remote sensing images have a wide coverage to generate the first inverted water quality data. On the other hand, the high-resolution remote sensing images have high resolution and rich data sources to obtain the second inverted water quality data. For small and medium-sized rivers where field sampling cannot be frequently carried out or is difficult to be carried out on site, and where monitoring stations are not installed, limited data can still be used to invert the water quality of the entire river. Water quality monitoring of complex rivers does not need to rely too much on manual labor, which reduces labor costs. At the same time, the monitoring cycle is flexible, and water quality inversion can be carried out in real time and efficiently to obtain dynamic water quality monitoring results. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0053] Figure 1 A schematic diagram of a flow chart of a river water quality inversion method based on remote sensing images in an embodiment of the present application;
[0054] Figure 2 A schematic diagram of the correspondence between a first reflectivity value of a first-resolution remote sensing image and measured water quality data in a river water quality inversion method based on remote sensing images in an embodiment of the present application;
[0055] Figure 3A schematic diagram of the correspondence between a second resolution value of a second resolution remote sensing image and first inverted water quality data in a river water quality inversion method based on remote sensing images in an embodiment of the present application;
[0056] Figure 4 Another schematic diagram of a flow chart of a river water quality inversion method based on remote sensing images in an embodiment of the present application;
[0057] Figure 5 This is a schematic diagram of the structure of a river water quality inversion device based on remote sensing images in an embodiment of the present application. DETAILED DESCRIPTION
[0058] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.
[0059] The embodiments of the present application provide a river water quality inversion method, equipment and medium based on remote sensing images, which are used to solve the current problem that water quality monitoring of complex rivers is overly dependent on manual labor, which consumes manpower costs, has a long monitoring cycle, and is difficult to obtain dynamic water quality monitoring results in real time and efficiently.
[0060] The following describes in detail various embodiments of the present application in conjunction with the accompanying drawings.
[0061] The present application embodiment provides a river water quality inversion method based on remote sensing images, such as Figure 1 As shown, the method may include steps S101-S105:
[0062] S101, the server determines first to-be-determined band combinations generated by a first predetermined number of remote sensing image bands corresponding to the same river monitoring point based on river monitoring information in a water quality monitoring work area.
[0063] The first undetermined band combination corresponds to a remote sensing image with a first resolution.
[0064] It should be noted that the server, as the executor of the river water quality inversion method based on remote sensing images, is only an example. The executor is not limited to the server, and this application does not make any specific limitations on this.
[0065] In the embodiment of the present application, before determining the first undetermined band combinations generated by the first predetermined number of remote sensing image bands corresponding to the same river monitoring point based on the river monitoring information in the water quality monitoring work area, the method further includes:
[0066] The server obtains the work area collection data from the user terminal. The work area collection data at least includes remote sensing image data, water quality measured data, and river system vector data in the water quality monitoring work area. The work area collection data is preprocessed to obtain river monitoring information. Among them, the river monitoring information at least includes a first resolution remote sensing image, a second resolution remote sensing image, and water quality measured data of multiple river monitoring points in the water quality monitoring work area. The water quality measured data corresponds to one or more water quality indicator dimensions. The resolution of the first resolution remote sensing image is smaller than that of the second resolution remote sensing image.
[0067] Specifically, the data collected in the work area may at least include remote sensing image data, measured water quality data, and river system vector data in the water quality monitoring work area. The water quality monitoring work area can be understood as the river area where the user wants to monitor the water quality, and the river area can especially be a small or medium-sized river. The remote sensing image data includes a first-resolution remote sensing image and a second-resolution remote sensing image. The first-resolution remote sensing image can be understood as a medium-resolution remote sensing image specified by the user, and the second-resolution remote sensing image is a high-resolution remote sensing image specified by the user. The medium-resolution and high-resolution division rules are set by the user according to actual use, and the resolution of the second-resolution remote sensing image is at least greater than the resolution of the first-resolution remote sensing image.
[0068] In actual use, the first-resolution remote sensing image can be crawled from a satellite image address survey website, and the second-resolution remote sensing image can be purchased from a satellite data supplier. For example, the first-resolution remote sensing image uses Sentinel-2 satellite images with a resolution of 10m; the second-resolution remote sensing image uses JL-01 satellite images with a resolution of 0.8m. . The measured water quality data can be some measured water quality data for river monitoring points such as national test sections, provincial test sections, and local monitoring stations, which can include permanganate index, water temperature, pH, dissolved oxygen concentration, conductivity, turbidity, ammonia nitrogen index, total phosphorus content, total nitrogen content, etc. River water system vector data can be understood as the boundary vector data of the water body.
[0069] The user terminal can be understood as a device such as a mobile phone or computer of a user who uploads the data collected in the work area, and this application does not make specific restrictions on this. The preprocessing of the data collected in the work area may include remote sensing image preprocessing, water quality measured data preprocessing, and river system vector data correction. Specifically, remote sensing image preprocessing includes performing radiation calibration, atmospheric correction, orthorectification, image fusion, image mosaicking and other processing work on the first resolution remote sensing image and the second resolution remote sensing image respectively; then, the two images processed in the previous step are geometrically aligned to form an aligned medium- and high-resolution remote sensing image orthophoto map, which is added to the river monitoring information. The preprocessing of measured water quality data, such as the monitoring data synchronized with the imaging time of the above-mentioned images according to the longitude and latitude of the section, the dimensions of the water quality indicators, and the imaging time of the above-mentioned images, is organized in a certain way to prepare for the subsequent water quality inversion work. River system vector data correction: For example, using ArcGIS software, the satellite image orthophoto map and river system vector data are superimposed, and the vector part that does not fit the water body boundary in the image is corrected, so that the river system vector is completely consistent with the water body boundary of the current remote sensing image.
[0070] In addition, river monitoring information can also include point source and surface source data and land use status data in the water quality monitoring work area after data cleaning. Point source and surface source data can be understood as point information data of industrial and mining enterprises, farmland, livestock breeding, etc.; land use status data refers to the status data of various land use types, including land use type, area, location and other information, so as to better display more information in the water quality monitoring distribution map.
[0071] In the embodiment of the present application, the first undetermined band combinations generated by determining the first predetermined number of remote sensing image bands corresponding to the same river monitoring point based on the river monitoring information in the water quality monitoring work area specifically include:
[0072] The server determines the first-resolution remote sensing image of the same river monitoring point according to the river monitoring information, selects a first predetermined number of remote sensing image bands from the remote sensing image bands corresponding to the first-resolution remote sensing image, and randomly performs four arithmetic operations on the selected remote sensing image bands to form corresponding first undetermined band combinations, until all remote sensing image bands corresponding to the first-resolution remote sensing image are randomly combined to determine each first undetermined band combination.
[0073] That is to say, this application will process the first-resolution remote sensing image and extract each remote sensing image band and its first reflectivity value of the first-resolution remote sensing image of the same river monitoring point, such as Figure 2As shown, the designated water quality indicator dimension of water quality monitoring is the permanganate index, and the remote sensing image bands corresponding to the first resolution remote sensing image of each river monitoring point include B1-B12, 12 remote sensing image bands. The first predetermined number of this application is set to 4, for example, that is, four remote sensing image bands form the first to-be-determined band combination, and the first predetermined number is set by the user according to the actual scenario, and this application does not make specific restrictions on this.
[0074] After selecting the first predetermined number of remote sensing image bands, the server can perform random arithmetic operations of "addition, subtraction, multiplication, and division" to combine the remote sensing image bands to obtain the first undetermined band combination, until all possible combinations are listed to obtain the first undetermined band combinations corresponding to each river monitoring point. For example, B6*(B11+B12) / B2 is one of the first undetermined band combinations.
[0075] S102, the server performs a correlation analysis on the first reflectivity value sets of each first pending band combination and the corresponding water quality measured data, so as to determine a first selected band combination corresponding to the specified water quality indicator dimension according to the correlation analysis result.
[0076] In the embodiment of the present application, a correlation analysis is performed on the first reflectivity value set of each first undetermined band combination and the corresponding water quality measured data, so as to determine the first selected band combination corresponding to the specified water quality indicator dimension according to the correlation analysis result, specifically including:
[0077] The server calculates the corresponding reflectance values of each first combination according to the first reflectance value sets corresponding to each first undetermined band combination. The first combination reflectance values and the water quality measured data of the corresponding specified water quality indicator dimension are respectively input into the preset Pearson correlation coefficient calculation formula to perform correlation analysis to obtain each first correlation coefficient. The first correlation coefficients are compared to determine that the first undetermined band combination corresponding to the maximum first correlation coefficient is the first selected band combination.
[0078] Among them, the server can calculate each first reflectivity value in the first reflectivity value set according to the first reflectivity value set and the four arithmetic operations combination corresponding to the first undetermined band combination, and use the calculation result as the first combination reflectivity value.
[0079] In other words, the server can obtain a first reflectivity value set consisting of the first reflectivity values corresponding to the first pending band combination, and use the four arithmetic operations corresponding to the first pending band combination, such as B6*(B11+B12) / B2 to calculate a final reflectivity value as the first combination reflectivity value. Subsequently, the server can input all the first combination reflectivity values and their corresponding water quality measured data into the Pearson correlation coefficient calculation formula for Pearson correlation analysis to obtain the first correlation coefficients between the first combination reflectivity values and the water quality measured data corresponding to various first pending band combinations. For the same first pending band combination of different river monitoring points, the present application can calculate the corresponding average values of the first correlation coefficients after obtaining the corresponding first correlation coefficients for correlation analysis. Therefore, the first correlation coefficients between different first to-be-determined band combinations are compared in size, the maximum value (maximum first correlation coefficient) among the first correlation coefficients corresponding to different first to-be-determined band combinations is determined, and the first to-be-determined band combination corresponding to the maximum first correlation coefficient is taken as the first selected band combination. For example, the first correlation coefficient of the first to-be-determined band combination B6*(B11+B12) / B2 is 0.9152, which is the maximum value among the first correlation coefficients, then the first selected band combination is B6*(B11+B12) / B2.
[0080] The Pearson correlation coefficient calculation formula is as follows:
[0081] r=∑[(xx mean )(yy mean )] / [sqrt(∑(xx mean ) 2 )*sqrt(∑(yy mean ) 2 )]
[0082] Among them, r is the correlation coefficient, which is between (-1, 1). The larger the value, the higher the correlation, such as the first correlation coefficient and the second correlation coefficient; x is the combined reflectivity value, such as the first combined reflectivity value and the second combined reflectivity value; y is the water quality monitoring data, x mean is the average value of the combined reflectance value, y mean It is the average value of water quality monitoring data.
[0083] In addition, this application can also be applied by Figure 2 Some of the measured data in are used as verification data, such as the data in sequence number 1-7 are used to obtain the above-mentioned first selected band combination, and the data in sequence number 8 is used to verify the accuracy of the above-mentioned first selected band combination.
[0084] In one embodiment of the present application, there are multiple water quality indicator dimensions when performing water quality monitoring, and there may be different first selected band combinations for different water quality indicator dimensions. Therefore, the present application can obtain the corresponding specified water quality indicator dimensions based on the user's specified operation on the user terminal. The specified water quality indicator dimensions include at least: permanganate index, water temperature, pH, dissolved oxygen concentration, conductivity, turbidity, ammonia nitrogen index, total phosphorus content, and total nitrogen content.
[0085] That is, there are different water quality indicator dimensions. Figure 2 The list shown is used to generate the first selected band combination and perform the subsequent corresponding water quality inversion steps.
[0086] S103, the server generates a first water quality monitoring inversion model based on the first selected band combination and the measured water quality data corresponding to the specified water quality indicator dimension, so as to determine a number of working area points and their corresponding first inverted water quality data according to the first water quality monitoring inversion model and the first resolution remote sensing image.
[0087] In the embodiment of the present application, based on the first selected band combination and the measured water quality data corresponding to the specified water quality indicator dimension, a first water quality monitoring inversion model is generated, specifically including:
[0088] First, the server determines each first selected band combination corresponding to a plurality of different river monitoring points and each corresponding first combination reflectance value. Then, the server determines each water quality measured data corresponding to the specified water quality indicator dimension obtained at the corresponding plurality of different river monitoring points. Then, a linear regression is performed on each first combination reflectance value and each water quality measured data to obtain a first water quality monitoring inversion model.
[0089] That is to say, when establishing the first water quality monitoring inversion model, the first combined reflectivity values and measured water quality data of different river monitoring points can be used for linear regression until the coefficient is determined to be greater than the first preset threshold, and the first water quality monitoring inversion model is established. The first preset threshold can be set by the user according to the actual usage scenario. The first water quality monitoring inversion model is, for example, Y1=a*X1+b, where Y1 is the value corresponding to the first inverted water quality data, a and b are the first coefficient and the second coefficient obtained by linear regression, and X1 is the first combined reflectivity value of each input work area point.
[0090] In the embodiment of the present application, a plurality of working area points and their corresponding first inverted water quality data are determined according to the first water quality monitoring inversion model and the first resolution remote sensing image, specifically including:
[0091] Based on the user's point designation operation on the first resolution remote sensing image through the user terminal, a number of corresponding working area points and their corresponding first combined reflectivity values are determined. The first combined reflectivity values corresponding to the working area points are respectively input into the first water quality monitoring inversion model, so that the output values of the first water quality monitoring inversion model are used as the first inverted water quality data of the corresponding working area points.
[0092] That is, the working area point can be specified by the user, the working area point has at least a first resolution remote sensing image, and the corresponding first combined reflectivity value can be calculated through the above step S102. After the first combined reflectivity value is input into the above first water quality monitoring inversion model, the data value of the first inverted water quality data corresponding to each working area point is obtained.
[0093] S104, the server generates a second selected band combination based on the second pending band combination corresponding to the river monitoring information and the first inverted water quality data, so as to generate a second water quality monitoring inversion model according to the second selected band combination and the first inverted water quality data.
[0094] The second undetermined band combination includes a combination of different remote sensing image bands corresponding to the second resolution remote sensing image.
[0095] That is to say, the second undetermined band combination is composed of the remote sensing image band combination of the high-resolution and second-resolution remote sensing images.
[0096] In an embodiment of the present application, based on the second undetermined band combination corresponding to the river monitoring information and the first inverted water quality data, a second selected band combination is generated, so as to generate a second water quality monitoring inversion model according to the second selected band combination and the first inverted water quality data, specifically including:
[0097] First, the server randomly combines the different remote sensing image bands corresponding to the second resolution remote sensing image in the river monitoring information to obtain the second undetermined band combinations. Then, the second combination reflectance values of the second undetermined band combinations and the corresponding first inverted water quality data are subjected to correlation analysis to generate the corresponding second selected band combinations according to the correlation analysis results. Then, the second combination reflectance values corresponding to the second selected band combinations and the corresponding first inverted water quality data are subjected to linear regression to obtain the second water quality monitoring inversion model according to the linear regression equation.
[0098] In other words, the second-resolution remote sensing image has multiple different remote sensing image bands. Taking the remote sensing image bands of JL-01 image as an example, Figure 3As shown, it includes four remote sensing image bands B1-B4. The user can select a second predetermined number such as 4, and randomly combine the four remote sensing image bands through four arithmetic operations to obtain a second pending band combination. The technical solution for the first pending band combination in step S101 may be referred to, which will not be described in detail here. After the server obtains the second pending band combination, it can also perform Pearson correlation analysis on the first inverted water quality data and the second combined reflectivity value of the working area point corresponding to the first inverted water quality data, such as calculating the second combined reflectivity value by (B1+B3) / (B1-B4). Subsequently, the second correlation coefficients corresponding to the obtained working area points are compared, and the second pending band combination corresponding to the maximum second correlation coefficient is used as the second selected band combination. The specific process can refer to the above-mentioned step S102, which will not be described in detail here. For example, if used Figure 3 Part of the data in the list is used to generate the second selected band combination, and the other part is used for accuracy verification.
[0099] The server will also use the second combined reflectivity value corresponding to the second selected band combination and the first inverted water quality data of the corresponding work area point for linear regression until its determination coefficient is greater than the second preset threshold, thereby obtaining the second water quality monitoring inversion model Y2=c*X2+d, where Y2 is the value of the second inverted water quality data, c and d are the third and fourth coefficients of the linear regression, and X2 is the second combined reflectivity value of the input second resolution remote sensing image. The second preset threshold can be set by the user according to the actual usage scenario, and this application does not make specific limitations on this.
[0100] Through the above scheme, this application combines medium and high resolution remote sensing images so that water quality inversion can be applicable to different types of rivers. For example, water quality can be deduced for small and medium-sized rivers through the method of this application without relying on on-site sampling or setting up cross-section monitoring stations on site.
[0101] S105, the server determines the second inverted water quality data corresponding to the water quality monitoring work area according to the second resolution remote sensing image and the second water quality monitoring inversion model, and adds the second inverted water quality data to the water quality monitoring distribution map, and displays it on the user terminal.
[0102] After obtaining the second water quality monitoring inversion model, the server can first determine the second combined reflectivity value of the second resolution remote sensing image at each location in the water quality monitoring work area, and input the second water quality monitoring inversion model to calculate the corresponding value of the second inverted water quality data at each location, and add it to the water quality monitoring distribution map and display it on the user terminal. It can be understood that the second inverted water quality data is displayed in different levels of color according to the numerical value, and the second inverted water quality data is displayed in the water quality monitoring work area of the water quality monitoring distribution map, that is, a river, so that the user can obtain the water quality monitoring results of the corresponding specified water quality indicator dimension according to the displayed content.
[0103] In addition, this application can also display the point source and surface source and land use status data of the work area on the map to better understand the river pollution situation.
[0104] Figure 4 Another schematic diagram of a flow chart of a river water quality inversion method based on remote sensing images provided in an embodiment of the present application is as follows: Figure 4 As shown, it includes: basic data collection, including remote sensing image data, ground water quality measured data, and other related data; remote sensing image data include medium-resolution remote sensing images, high-resolution remote sensing images, and data elevation models; ground water quality measured data include national section monitoring data, provincial section monitoring data, and local monitoring station data; other related data include river system vector data, point source and surface source data, and land use status data; data preprocessing, including the matched medium- and high-resolution remote sensing image orthophoto base map, the monitoring data of the watershed section of the working area matched with the remote sensing image, and the point source and surface source and land use status data of the working area; remote sensing inversion of water quality in the watershed of the working area based on medium resolution, including the monitoring data of the watershed section location and the corresponding medium-resolution image, after pre-processing The medium-resolution orthophotos, cross-section monitoring data and corresponding image reflectance values, the best band combination of remote sensing images is determined, the remote sensing inversion model of water quality in the basin is established, the remote sensing inversion of water quality in the basin is carried out, the accuracy of the inversion results is verified, and the measured data of ground water quality in the working area is generated; the remote sensing inversion of water quality in the working area based on high resolution includes the measured data of ground water quality generated in the previous step, the preprocessed high-resolution orthophotos, the measured data and the corresponding image reflectance values are generated, the best band combination of remote sensing images is determined, the remote sensing inversion model of water quality in the working area is established, the remote sensing inversion of water quality in the working area is carried out, the accuracy of the inversion results is verified, and the water quality status monitoring results of the working area; the current data of point sources, surface sources and land use in the working area; the spatial distribution characteristics and analysis of the water quality status of small and medium-sized rivers and their small tributaries.
[0105] Through the above technical solution, the present application can use two types of resolution remote sensing images in the water quality monitoring work area to respectively establish water quality monitoring inversion models. On the one hand, the medium-resolution remote sensing images have a wide coverage to generate the first inverted water quality data. On the other hand, the high-resolution remote sensing images have high resolution and rich data sources to obtain the second inverted water quality data. For small and medium-sized rivers where field sampling cannot be frequently carried out or is difficult to be carried out on site, and where monitoring stations are not installed, limited data can still be used to invert the water quality of the entire river. Water quality monitoring of complex rivers does not need to rely too much on manual labor, which reduces labor costs. At the same time, the monitoring cycle is flexible, and water quality inversion can be carried out in real time and efficiently to obtain dynamic water quality monitoring results.
[0106] Figure 5A schematic diagram of the structure of a river water quality inversion device based on remote sensing images provided in an embodiment of the present application is shown in FIG. Figure 5 As shown, the device includes:
[0107] At least one processor; and a memory in communication with the at least one processor. The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to:
[0108] Based on the river monitoring information in the water quality monitoring work area, determine the first predetermined number of remote sensing image bands corresponding to the same river monitoring point to generate each first undetermined band combination. Wherein, the first undetermined band combination corresponds to the first resolution remote sensing image. Perform correlation analysis on the first reflectivity value set of each first undetermined band combination and the corresponding water quality measured data, so as to determine the first selected band combination corresponding to the specified water quality index dimension according to the correlation analysis result. Based on the first selected band combination and the water quality measured data corresponding to the specified water quality index dimension, generate a first water quality monitoring inversion model, so as to determine a number of work area points and their corresponding first inversion water quality data according to the first water quality monitoring inversion model and the first resolution remote sensing image. Based on the second undetermined band combination corresponding to the river monitoring information and the first inversion water quality data, generate a second selected band combination, so as to generate a second water quality monitoring inversion model according to the second selected band combination and the first inversion water quality data. Wherein, the second undetermined band combination includes a combination of different remote sensing image bands corresponding to the second resolution remote sensing image. The second resolution remote sensing image is input into the second water quality monitoring inversion model, the second inverted water quality data corresponding to the water quality monitoring work area is determined, and the second inverted water quality data is added to the water quality monitoring distribution map and displayed on the user terminal.
[0109] The embodiment of the present application further provides a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured as follows:
[0110] Based on the river monitoring information in the water quality monitoring work area, determine the first predetermined number of remote sensing image bands corresponding to the same river monitoring point to generate each first undetermined band combination. Wherein, the first undetermined band combination corresponds to the first resolution remote sensing image. Perform correlation analysis on the first reflectivity value set of each first undetermined band combination and the corresponding water quality measured data, so as to determine the first selected band combination corresponding to the specified water quality index dimension according to the correlation analysis result. Based on the first selected band combination and the water quality measured data corresponding to the specified water quality index dimension, generate a first water quality monitoring inversion model, so as to determine a number of work area points and their corresponding first inversion water quality data according to the first water quality monitoring inversion model and the first resolution remote sensing image. Based on the second undetermined band combination corresponding to the river monitoring information and the first inversion water quality data, generate a second selected band combination, so as to generate a second water quality monitoring inversion model according to the second selected band combination and the first inversion water quality data. Wherein, the second undetermined band combination includes a combination of different remote sensing image bands corresponding to the second resolution remote sensing image. The second resolution remote sensing image is input into the second water quality monitoring inversion model, the second inverted water quality data corresponding to the water quality monitoring work area is determined, and the second inverted water quality data is added to the water quality monitoring distribution map and displayed on the user terminal.
[0111] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.
[0112] The devices and media provided in the embodiments of the present application correspond one-to-one to the methods. Therefore, the devices and media also have similar beneficial technical effects as the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.
[0113] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0114] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.
Claims
1. A river water quality inversion method based on remote sensing images, It is characterized in that The method comprises: Based on the river monitoring information in the water quality monitoring work area, determine each first undetermined band combination generated by a first predetermined number of remote sensing image bands corresponding to the same river monitoring point; wherein the first undetermined band combination corresponds to a first resolution remote sensing image; Performing a correlation analysis on the first reflectivity value sets of each of the first undetermined band combinations and the corresponding measured water quality data, so as to determine a first selected band combination corresponding to a specified water quality index dimension according to the correlation analysis result; the specified water quality index dimension includes at least: permanganate index, water temperature, pH, dissolved oxygen concentration, conductivity, turbidity, ammonia nitrogen index, total phosphorus content, and total nitrogen content; Based on the first selected band combination and the measured water quality data corresponding to the specified water quality indicator dimension, a first water quality monitoring inversion model is generated to determine a plurality of working area points and their corresponding first inverted water quality data according to the first water quality monitoring inversion model and the first resolution remote sensing image; Based on the second undetermined band combination corresponding to the river monitoring information and the first inverted water quality data, a second selected band combination is generated to generate a second water quality monitoring inversion model according to the second selected band combination and the first inverted water quality data; wherein the second undetermined band combination includes a combination of different remote sensing image bands corresponding to the second resolution remote sensing image; According to the second resolution remote sensing image and the second water quality monitoring inversion model, the second inverted water quality data corresponding to the water quality monitoring work area is determined, and the second inverted water quality data is added to the water quality monitoring distribution map and displayed on the user terminal.
2. A river water quality inversion method based on remote sensing images according to claim 1, It is characterized in that Before determining first undetermined band combinations generated by a first predetermined number of remote sensing image bands corresponding to the same river monitoring point based on the river monitoring information in the water quality monitoring work area, the method further includes: Acquire work area collection data from the user terminal; the work area collection data at least includes remote sensing image data, water quality measured data, and river system vector data in the water quality monitoring work area; The data collected in the working area are preprocessed to obtain the river monitoring information; wherein the river monitoring information at least includes a first-resolution remote sensing image, a second-resolution remote sensing image, and the measured water quality data of multiple river monitoring points in the water quality monitoring working area; the measured water quality data corresponds to one or more water quality indicator dimensions; the resolution of the first-resolution remote sensing image is smaller than that of the second-resolution remote sensing image.
3. The river water quality inversion method based on remote sensing images according to claim 1, It is characterized in that Based on the river monitoring information in the water quality monitoring work area, determining each first undetermined band combination generated by a first predetermined number of remote sensing image bands corresponding to the same river monitoring point specifically includes: Determine the first-resolution remote sensing image of the same river monitoring point according to the river monitoring information; The first predetermined number of remote sensing image bands are screened from the remote sensing image bands corresponding to the first resolution remote sensing image, and four arithmetic operations are randomly performed on the screened remote sensing image bands to form corresponding first pending band combinations, until all the remote sensing image bands corresponding to the first resolution remote sensing image are randomly combined to determine the first pending band combinations.
4. The method for river water quality inversion based on remote sensing images according to claim 1, It is characterized in that Performing a correlation analysis on the first reflectivity value sets of each of the first undetermined band combinations and the corresponding measured water quality data, so as to determine the first selected band combination corresponding to the specified water quality indicator dimension according to the correlation analysis result, specifically including: Calculate the corresponding reflectance values of each first combination according to the first reflectance value sets respectively corresponding to the first undetermined waveband combinations; Input each of the first combined reflectance values and the water quality measured data of the designated water quality index dimension corresponding thereto into a preset Pearson correlation coefficient calculation formula to perform correlation analysis to obtain each first correlation coefficient; The first correlation coefficients are compared to determine that the first undetermined band combination corresponding to the maximum first correlation coefficient is the first selected band combination.
5. A river water quality inversion method based on remote sensing images according to claim 4, It is characterized in that The method further comprises: Based on the operation specified by the user on the user terminal, the corresponding specified water quality indicator dimension is obtained.
6. A river water quality inversion method based on remote sensing images according to claim 4, It is characterized in that Based on the first selected band combination and the measured water quality data corresponding to the specified water quality indicator dimension, generating a first water quality monitoring inversion model specifically includes: Determine each of the first selected band combinations corresponding to a plurality of different river monitoring points and each of the first combination reflectance values corresponding to the combinations; Determine each of the water quality measured data corresponding to the specified water quality indicator dimension obtained at the corresponding plurality of different river monitoring points; Performing linear regression on each of the first combined reflectivity values and each of the water quality measured data to obtain the first water quality monitoring inversion model; Determining a plurality of working area points and their corresponding first inverted water quality data according to the first water quality monitoring inversion model and the first resolution remote sensing image, specifically including: Based on a point designation operation performed by a user through a user terminal on the remote sensing image of the first resolution, determining a corresponding number of the working area points and their corresponding first combined reflectivity values; The first combined reflectivity values corresponding to each of the working area points are respectively input into the first water quality monitoring inversion model, so that the output values of the first water quality monitoring inversion model are used as the first inverted water quality data of the corresponding each of the working area points.
7. A river water quality inversion method based on remote sensing images according to claim 4, It is characterized in that Calculating the corresponding reflectance values of each first combination according to the first reflectance value sets respectively corresponding to the first undetermined waveband combinations specifically includes: According to the first reflectivity value set and four arithmetic operations combination corresponding to the first undetermined band combination, each first reflectivity value in the first reflectivity value set is calculated according to the four arithmetic operations combination, so as to use the calculation result as the first combination reflectivity value.
8. The method for river water quality inversion based on remote sensing images according to claim 1, It is characterized in that Based on the second undetermined band combination corresponding to the river monitoring information and the first inverted water quality data, a second selected band combination is generated, so as to generate a second water quality monitoring inversion model according to the second selected band combination and the first inverted water quality data, specifically including: Randomly combining different remote sensing image bands corresponding to the remote sensing image with the second resolution in the river monitoring information to obtain the second undetermined band combinations; Performing correlation analysis on the second combined reflectivity values of each of the second undetermined band combinations and the first inverted water quality data respectively corresponding thereto, so as to generate the corresponding second selected band combination according to the correlation analysis result; A linear regression is performed on the second combined reflectivity values corresponding to each of the second selected band combinations and the corresponding first inverted water quality data to obtain the second water quality monitoring inversion model according to the linear regression equation.
9. A river water quality inversion device based on remote sensing images, It is characterized in that The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a river water quality inversion method based on remote sensing images as described in any one of claims 1-8 above.
10. A non-volatile computer storage medium storing computer executable instructions, It is characterized in that The computer executable instructions can execute a river water quality inversion method based on remote sensing images as described in any one of claims 1 to 8 above.
