River flow remote sensing estimation method, electronic equipment and storage medium
By constructing a reflectivity flow inversion model based on river width, the problem of large differences in the remote sensing estimate value of the river flow and the actual measured value is solved, and accurate estimation and efficient monitoring of river flow are achieved.
Patent Information
- Application Number
- CN202410949690.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-16
- Publication Date
- 2024-11-08
AI Technical Summary
In the prior art, the estimated value of the remote sensing estimation method of river flow is quite different from the measured value, and is greatly affected by river morphology and environmental factors, making it difficult to accurately monitor and predict.
By obtaining river remote sensing time series data for preprocessing, water body information in the region of interest is extracted, river width information is calculated, and a reflectivity flow inversion model that takes into account the river width is constructed based on the reflectivity of calibration pixels, permanent water body pixels and measured pixels. , conduct remote sensing estimation of river flow.
It realizes a relatively accurate estimation of river flow data, has small system errors, simple and efficient algorithms, and can obtain higher river flow data under the conditions of minimal known data.
Smart Images

Figure CN118918467A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a method for estimating river flow data, and in particular to a river flow remote sensing estimation method, electronic equipment and storage medium. Background Art
[0002] Water resources are a key factor in maintaining ecosystem stability and promoting sustainable social development. Globally, accurate monitoring and prediction of river runoff data are crucial for water resource management and flood and drought disaster warning.
[0003] With the continuous improvement of remote sensing technology and the prediction ability of existing methods, the use of remote sensing technology for river flow inversion has great potential. However, there are still a series of problems in current research. For example, in the methods of estimating river flow by remote sensing, many methods require some known measured flow data as an aid, and field hydrological monitoring usually requires huge human, material and financial resources. The limited measured data brings many inconveniences to the research. Secondly, due to the limitations of environmental and economic factors in the area where the river is located, it is impossible to establish measured hydrological stations in many areas. At the same time, the flow of rivers is also affected by many factors, including precipitation, evaporation, basin topography, etc. In the method of estimating river flow using remote sensing images, the quality of remote sensing images also directly affects the accuracy of the estimation results. Therefore, the accuracy of river flow estimation based solely on remote sensing images is often limited, and the estimated flow value is quite different from the measured value. Summary of the invention
[0004] The purpose of the present invention is to provide a river flow remote sensing estimation method, electronic equipment and storage medium to solve the problem that the estimated value of river flow in the prior art is quite different from the measured value.
[0005] To achieve one of the above-mentioned purposes, an embodiment of the present invention provides a method for remote sensing estimation of river flow, the method comprising:
[0006] Obtain river remote sensing time series data for preprocessing and select a region of interest of a preset size;
[0007] Extracting water body information in the region of interest, and calculating river width information in the water body information;
[0008] Respectively obtaining the reflectance of the calibration pixels, permanent water pixels and measurement pixels in the region of interest;
[0009] A reflectivity flow inversion model taking into account the river width is constructed based on the river width information and the reflectivity of the calibration pixels, permanent water body pixels and measurement pixels, and remote sensing estimation of river flow is performed.
[0010] As a further improvement of an embodiment of the present invention, the method further includes that the pretreatment includes:
[0011] The data with cloud content greater than a preset value in the time series data of river remote sensing are discarded.
[0012] As a further improvement of an embodiment of the present invention, the method further includes that extracting water body information in the region of interest includes:
[0013] The normalized difference water index is used to enhance water body characteristics, and water body information and non-water body information in the region of interest are extracted.
[0014] As a further improvement of an embodiment of the present invention, the method further includes that calculating the river width information in the water body information includes:
[0015] When the river section where the water body information is located is a multi-branch river, the river width information is calculated using the average width method;
[0016] When the river section where the water body information is located is a single-shaped river channel, the orthogonal skeleton method is used to calculate the river width information.
[0017] As a further improvement of an embodiment of the present invention, the method further includes that obtaining the reflectivity of the calibration pixel of the region of interest includes:
[0018] Extracting the reflectivity of each pixel point in the region of interest in the time series;
[0019] For each pixel point, the mean value mean1 and the standard deviation SD1 of the reflectivity in the time series are calculated;
[0020] By calculating the coefficient of variation CV, selecting pixel points whose coefficient of variation is less than a preset value, and obtaining the average reflectivity of the pixel points whose coefficient of variation is less than the preset value as the reflectivity C of the calibration pixel;
[0021] The calculation formula of the coefficient of variation CV is expressed as:
[0022]
[0023] As a further improvement of an embodiment of the present invention, the method further includes obtaining the reflectivity of the permanent water body pixels in the region of interest, including:
[0024] Expanding the water body information outwards by a buffer zone of a preset size to produce a water mask, and selecting and marking all pixels within the water mask;
[0025] For each pixel in the water mask, calculate the mean value mean2 and standard deviation SD2 of its reflectivity in the time series;
[0026] Pixel points whose mean2*SD2 is less than a preset threshold are selected, and the average reflectivity of the pixel points whose mean2*SD2 is less than the preset threshold is calculated as the reflectivity W of the permanent water body pixel.
[0027] As a further improvement of an embodiment of the present invention, the method further includes that obtaining the measured pixel information of the region of interest includes:
[0028] Loop through all pixels of the region of interest in the preprocessed time series data to obtain the reflectivity of each pixel in the time series;
[0029] Calculate the Spearman correlation MC between the reflectance of each pixel in the time series and the reflectance of the calibration pixel, and the Spearman correlation MW between the reflectance of each pixel and the permanent water body pixel;
[0030] Pixel points whose reflectances are smaller than both the MC and MW preset thresholds are selected, and the average reflectance of the pixel points whose reflectances are smaller than both the MC and MW preset thresholds is calculated as the reflectance M of the measured pixel.
[0031] As a further improvement of an embodiment of the present invention, the method further includes that the reflectivity flow inversion model taking into account the river width is expressed as:
[0032]
[0033] Among them, Q is the river flow; Width is the river width; z, k, and b are coefficients.
[0034] To achieve one of the above-mentioned purposes of the invention, an embodiment of the present invention further provides an electronic device, comprising a memory and a processor, characterized in that the memory stores a computer program that can be run on the processor, and when the program is executed on the processor, the steps in the above-mentioned river flow remote sensing estimation method are implemented.
[0035] To achieve one of the above-mentioned objects of the invention, an embodiment of the present invention further provides a storage medium, wherein the storage medium stores a computer program, and is characterized in that when the computer program is executed by a processor, the steps in the above-mentioned river flow remote sensing estimation method are implemented.
[0036] Compared with the prior art, the present invention provides a river flow remote sensing estimation method, electronic device and storage medium. By introducing width changes on the basis of the CMW method, the influence of river morphology on river flow is taken into account in the CMW method, so that river flow data can be obtained more accurately, and the systematic error between the estimated value and the true value is small, the algorithm is simple and the efficiency is high. Relatively high river flow data can be obtained using the least known data. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Figure 1 It is an overall flow chart of the river flow remote sensing estimation method of the present invention.
[0038] Figure 2 It is a river flow estimation comparison diagram of an embodiment of the river flow remote sensing estimation method of the present invention. DETAILED DESCRIPTION
[0039] The present invention will be described in detail below in conjunction with the specific embodiments shown in the accompanying drawings. However, these embodiments do not limit the present invention, and any structural, methodological, or functional changes made by a person skilled in the art based on these embodiments are all within the scope of protection of the present invention.
[0040] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.
[0041] In the first embodiment of the present invention, the present invention provides a method for estimating river flow remote sensing, such as Figure 1 As shown, the method includes,
[0042] S1: Obtain river remote sensing time series data for preprocessing and select a region of interest of a preset size;
[0043] S2: extracting water body information in the region of interest, and calculating river width information in the water body information;
[0044] S3: respectively obtaining the reflectance of the calibration pixels, permanent water pixels and measurement pixels in the region of interest;
[0045] S4: Based on the river width information and the reflectivity of the calibration pixels, permanent water body pixels and measurement pixels, a reflectivity flow inversion model that takes into account the river width is constructed to perform remote sensing estimation of river flow.
[0046] In a specific embodiment of the present invention, the pretreatment comprises:
[0047] The data with cloud content greater than a preset value in the time series data of river remote sensing are discarded.
[0048] Furthermore, the present invention selects near infrared band 8 (central wavelength 832.8nm) in Sentinel-2 multispectral image as the experimental band, and the spatial resolution of this band is 10m. For the historical remote sensing data of the river obtained, the images with cloud content greater than 20% are first discarded, and the pixel reflectance around the hydrological station is checked to see if it is reasonable and normal. Then, the specific location of the hydrological station is located according to the longitude and latitude. For the study area, the upstream or downstream river section near the hydrological station is selected from each image as the region of interest ROI, and the ROI size is J×K.
[0049] In a specific embodiment of the present invention, the water body information in the region of interest is extracted, specifically,
[0050] The normalized difference water index is used to enhance water body characteristics, and water body information and non-water body information in the region of interest are extracted.
[0051] Furthermore, the normalized difference water index NDWI is used to enhance the water body characteristics, and the maximum inter-class variance method is used to determine the optimal threshold of the NDWI image; the normalized difference water index NDWI is expressed as:
[0052]
[0053] Among them, ρ green , nir Represent the reflectance of the green band and the near-infrared band respectively.
[0054] It should be noted that NDWI makes full use of the characteristics of water body's strong absorption of near-infrared light and strong reflection of green light, and performs normalized difference processing on the near-infrared and green light bands, so that the water area is convex and the non-water area is suppressed. green , nir They correspond to the B3 and B8 bands in the Sentinel-2 data respectively.
[0055] In a specific embodiment of the present invention, the river width information in the water body information is calculated, specifically, when the river section where the water body information is located is a multi-branch river channel, the river width information is calculated using the average width method;
[0056] When the river section where the water body information is located is a single-shaped river channel, the orthogonal skeleton method is used to calculate the river width information.
[0057] It should be noted that the average width method is a method of obtaining the average river width by calculating the water body area and the length of the river section. This method is suitable for waters with small changes in river morphology. Specifically, after the water body is extracted, the water body area is calculated according to the pixel spatial resolution and the number of water body pixels in the target river section. Then the river centerline is extracted for the target river section in the study area, the length of the river section centerline is used as the river section length, and the average width of the river section is calculated by the algorithm formula. The orthogonal skeleton method is to calculate the width of the river section and take the average value of the width of multiple sections in the river section as the river section width. This method is more suitable for rivers with large changes in river morphology. Its basic process is to first extract the river centerline and edge line of the water body mask obtained by water body extraction, and then use the singular value decomposition (SVD) method to determine the normal vector for each point on the center line, establish a local coordinate system, and finally calculate the distance from the origin to the point on the edge line with the point on the center line as the origin. The sum of the distances is the required river section width.
[0058] In a specific embodiment of the present invention, the reflectivity of the calibration pixel of the region of interest is obtained, specifically,
[0059] Extract the reflectivity of each pixel in the region of interest within the time series;
[0060] For each pixel point, the mean value mean1 and the standard deviation SD1 of the reflectivity in the time series are calculated;
[0061] By calculating the coefficient of variation CV, selecting pixel points whose coefficient of variation is less than a preset value, and obtaining the average reflectivity of the pixel points whose coefficient of variation is less than the preset value as the reflectivity C of the calibration pixel;
[0062] The calculation formula of the coefficient of variation CV is expressed as:
[0063]
[0064] Furthermore, in the present invention, it is necessary to select several relatively stable pixels in the long-time series remote sensing images as calibration pixels. The coefficient of variation (CV), also known as the "discrete coefficient", is a normalized measure of the degree of discreteness of the probability distribution. It is defined as the ratio of the standard deviation to the mean value, which can reflect the stability of the remote sensing image reflectance in the long-time series of pixels at the same position. Therefore, the coefficient of variation CV is selected to determine the position of the calibration pixel. First, the time series of J×K pixels in the ROI is traversed in a loop to obtain J×K values, and then the J×K values are arranged in ascending order, and the pixel position corresponding to the value at the top 3 percentiles is selected as the specific position of the calibration pixel. The purpose of selecting pixels at multiple positions as calibration pixels is to avoid the uncertainty caused by the missing of a single pixel or the sudden change of the pixel reflectance. Finally, the average value of the reflectance of the pixels at the selected multiple positions is obtained as the reflectance of the calibration pixel.
[0065] In a specific embodiment of the present invention, the reflectivity of the permanent water pixel in the region of interest is obtained, specifically,
[0066] Expanding the water body information outwards by a buffer zone of a preset size to produce a water mask, and selecting and marking all pixels within the water mask;
[0067] For each pixel in the water mask, calculate the mean value mean2 and standard deviation SD2 of its reflectivity in the time series;
[0068] Pixel points whose mean2*SD2 is less than a preset threshold are selected, and the average reflectivity of the pixel points whose mean2*SD2 is less than the preset threshold is calculated as the reflectivity W of the permanent water body pixel.
[0069] Furthermore, using the "maximum range" data in the JRC global surface water map, a buffer of 3 pixels around each pixel is used to make a water mask. Only the pixels included in the water mask are considered, and the product between the time average of the reflectivity of each pixel (the lower, the greater the possibility of water) and its time standard deviation (the lower, the smaller the change in the pixel) is calculated. The lower the result value of the product, the more likely it is that the pixel at this location exists in the water body regardless of high or low flow. By setting a percentage threshold, pixels with a result value less than the threshold are identified as the location of permanent water pixels, and the average value of the pixels corresponding to the permanent water pixel location in each image is calculated to obtain the time series of permanent water pixels. The reflectivity of permanent water pixels represents the reflectivity of sediments.
[0070] Furthermore, in a specific implementation scenario, by setting a threshold of 0.05, pixels with result values less than the threshold are identified as the locations of permanent water pixels.
[0071] In a specific embodiment of the present invention, obtaining the measured pixel information of the region of interest includes:
[0072] Loop through all pixels of the region of interest in the preprocessed time series data to obtain the reflectivity of each pixel in the time series;
[0073] Calculate the Spearman correlation MC between the reflectance of each pixel in the time series and the reflectance of the calibration pixel, and the Spearman correlation MW between the reflectance of each pixel and the permanent water body pixel;
[0074] Pixel points whose reflectances are smaller than both the MC and MW preset thresholds are selected, and the average reflectance of the pixel points whose reflectances are smaller than both the MC and MW preset thresholds is calculated as the reflectance M of the measured pixel.
[0075] Furthermore, the location of the measurement pixel needs to reflect both the changes in high flow and low flow. Specifically, pixels that are sensitive to low flow should have poor correlation with calibration pixels (because they are flooded most of the time), while those that are sensitive to high flow should have poor correlation with permanent water pixels (because they are dry most of the time). Therefore, one of the conditions that a pixel must meet to be selected as the best measurement pixel is that its reflectivity has a very low Spearman correlation with permanent water pixels and calibration pixels. First, the time series of each pixel is obtained by looping through the J×K pixels of the multi-period image ROI, and then the Spearman correlation of the time series value with the permanent water pixel and the calibration pixel time series is calculated respectively, and the results are recorded as MW and MC respectively. The lower the correlation value, the more likely it is the best location for the measurement pixel. A certain threshold is set for MW and MC. When a pixel satisfies both of them, it means that it is the best location for the measurement pixel. In order to avoid selecting permanently flooded pixels (because they have poor correlation with calibration pixels), the threshold of the correlation between the measurement pixel and the permanent water pixel is generally larger than the threshold of the correlation between the calibration pixel.
[0076] It should be noted that the formula for calculating the Spearman correlation is:
[0077]
[0078] Among them, the original data x i With y i Arrange them in order from large to small, record R(x i ) is x i The ranking of size in x, R(y i ) is y i The size ranking in y, is the x-rank mean, is the mean of the y rank, and n is the number of data pairs.
[0079] Furthermore, in a specific implementation scenario, the thresholds of MW and MC are set to 0.6 and 0.4 respectively.
[0080] In one embodiment of the present invention, the reflectivity flow inversion model taking into account the river width (CMWW model) is expressed as:
[0081]
[0082] Among them, Q is the river flow; Width is the river width; z, k, and b are coefficients.
[0083] It should be noted that z is a numerical coefficient, z=max(MW)+min(M); k and b are coefficients to be solved, which are obtained by performing a function fitting with the known Q and the calculated C, W, M, and Width data.
[0084] Furthermore, after the positions of calibration pixels, permanent water body pixels, and measurement pixels are determined, the reflectivity ratio is calculated for each period. In order to reduce the impact of data mutations, an exponential filter is used here to first smooth the data to reduce the noise level of remote sensing image pixels, and finally the flow estimation is performed according to the CMWW model.
[0085] Furthermore, Figure 2 As shown in the figure, the river flow value estimated by the present invention is close to the trend of runoff change measured by the hydrological station, and the estimated result can better reflect the change of runoff. The estimated result is closer to the measured flow value, and the constructed model is more stable. At the same time, when the flow is large, the flow estimated by the CMWW method after fusion width is closer to the true value.
[0086] In a second embodiment of the present invention, the present invention provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the program is executed on the processor, the steps in the river flow remote sensing estimation method as described in any one of the above items are implemented.
[0087] In the third embodiment of the present invention, the present invention provides a storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps in any one of the above-mentioned river flow remote sensing estimation methods.
[0088] In summary, the present invention provides a river flow remote sensing estimation method, electronic device and storage medium. By introducing width changes on the basis of the CMW method, the influence of river morphology on river flow is taken into account in the CMW method, so that river flow data can be obtained more accurately, and the systematic error between the estimated value and the true value is small, the algorithm is simple and the efficiency is high. Relatively high river flow data can be obtained using the least known data.
[0089] It should be understood that although this specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each implementation mode may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.
[0090] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the modules described above can refer to the corresponding process in the aforementioned method implementation, and will not be repeated here.
[0091] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment.
[0092] In addition, each functional module in each embodiment of the present application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The above integrated modules can be implemented in the form of hardware or in the form of hardware plus software functional modules.
[0093] The above-mentioned integrated module implemented in the form of a software function module can be stored in a computer-readable storage medium. The above-mentioned software function module is stored in a storage medium, including a number of instructions for a computer system (which can be a personal computer, a server, or a network system, etc.) or a processor to perform some steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store program codes.
[0094] Finally, it should be noted that the above implementation modes are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned implementation modes, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned implementation modes, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various implementation modes of the present application.
Claims
1. A method for estimating river flow by remote sensing, characterized in that: include, Obtain river remote sensing time series data for preprocessing and select a region of interest of a preset size; Extracting water body information in the region of interest, and calculating river width information in the water body information; Respectively obtaining the reflectance of the calibration pixels, permanent water pixels and measurement pixels in the region of interest; A reflectivity flow inversion model taking into account the river width is constructed based on the river width information and the reflectivity of the calibration pixels, permanent water body pixels and measurement pixels, and remote sensing estimation of river flow is performed.
2. The method for estimating river flow by remote sensing according to claim 1, characterized in that: The pretreatment includes: The data with cloud content greater than a preset value in the time series data of river remote sensing are discarded.
3. The method for estimating river flow by remote sensing according to claim 2, characterized in that: The extracting of water body information in the region of interest comprises: The normalized difference water index is used to enhance water body characteristics, and water body information and non-water body information in the region of interest are extracted.
4. The method for estimating river flow by remote sensing according to claim 3, characterized in that: Calculating the river width information in the water body information includes: When the river section where the water body information is located is a multi-branch river, the river width information is calculated using the average width method; When the river section where the water body information is located is a single-shaped river channel, the orthogonal skeleton method is used to calculate the river width information.
5. The method for estimating river flow by remote sensing according to claim 2, characterized in that: Obtaining the reflectivity of the calibration pixel of the region of interest includes, Extracting the reflectivity of each pixel point in the region of interest in the time series; For each pixel point, the mean value mean1 and the standard deviation SD1 of the reflectivity in the time series are calculated; By calculating the coefficient of variation CV, selecting pixel points whose coefficient of variation is less than a preset value, and obtaining the average reflectivity of the pixel points whose coefficient of variation is less than the preset value as the reflectivity C of the calibration pixel; The calculation formula of the coefficient of variation CV is expressed as:
6. The method for estimating river flow by remote sensing according to claim 2, characterized in that: Obtaining the reflectance of the permanent water pixel of the region of interest includes, Expanding the water body information outwards by a buffer zone of a preset size to produce a water mask, and selecting and marking all pixels within the water mask; For each pixel in the water mask, calculate the mean value mean2 and standard deviation SD2 of its reflectivity in the time series; Pixel points whose mean2*SD2 is less than a preset threshold are selected, and the average reflectivity of the pixel points whose mean2*SD2 is less than the preset threshold is calculated as the reflectivity W of the permanent water body pixel.
7. The method for estimating river flow by remote sensing according to claim 2, characterized in that: Acquiring the measured pixel information of the region of interest includes: Loop through all pixels of the region of interest in the preprocessed time series data to obtain the reflectivity of each pixel in the time series; Calculate the Spearman correlation MC between the reflectance of each pixel in the time series and the reflectance of the calibration pixel, and the Spearman correlation MW between the reflectance of each pixel and the permanent water body pixel; Pixel points whose reflectances are smaller than both the MC and MW preset thresholds are selected, and the average reflectance of the pixel points whose reflectances are smaller than both the MC and MW preset thresholds is calculated as the reflectance M of the measured pixel.
8. The method for estimating river flow by remote sensing according to claim 1, characterized in that: The reflectivity flow inversion model taking into account the river width is expressed as: Among them, Q is the river flow; Width is the river width; z, k, and b are coefficients.
9. An electronic device, comprising a memory and a processor, characterized in that: The memory stores a computer program that can be run on the processor, and when the program is executed on the processor, the steps in the method for extracting remote sensing estimation of river flow as described in any one of claims 1-8 are implemented.
10. A storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps in the river flow remote sensing estimation method as claimed in any one of claims 1 to 8 are implemented.
Citation Information
Cited By
A multi-strategy, integrated, automated river channel width extraction method based on remote sensing data.
JP7933436B1
