Automatic riverway section width correction method based on B-G segmentation and M-K trend inspection

Abstract

The invention discloses an automatic riverway section width correction method based on B-G segmentation and M-K trend inspection. The method comprises the steps that the gradient and the slope direction of a research area are calculated; generating an initial river channel section; section information extraction; performing section gradient M-K trend inspection; performing section slope direction B-G sudden change inspection; and adjusting the section width according to the results of the slope M-K trend test and the slope direction B-G sudden change test. According to the method, the slope periodic change and the gradient trend change of the river channel are comprehensively considered, the morphological characteristics of the river channel section can be analyzed more accurately, automatic correction of the width of the river channel section is achieved, and the method has the advantages of being objective, convenient to implement, wide in application area and the like.

Description

Technical Field

[0001] The invention relates to the technical field of river section width, and in particular to an automatic correction method for river section width based on BG segmentation and MK trend test. Background Art

[0002] River channel topography data is an important basic data for constructing hydrological and hydrodynamic numerical models, and is also an important reference for studying the river channel evolution process and river channel engineering design. The width of the river channel is closely related to the river flow. Accurate river channel cross-section data can improve the accuracy of engineering design and flood simulation.

[0003] Traditional ground-based mapping methods can accurately reflect the actual situation of rivers, but because river section data is difficult to obtain or the measurement cost is too high, the current extraction of river terrain data is mainly through digital elevation models combined with visual interpretation of remote sensing images. With the development of computer technology and remote sensing technology, the temporal resolution and spatial resolution of remote sensing images have been significantly improved. Remote sensing image interpretation based on deep learning provides a new method for river section extraction and makes it possible to quickly identify complex boundaries of water areas. However, the morphology of the river in the image is related to the changes in the situation of water resources. For example, the construction of upstream reservoirs and floods and droughts will affect the water body identification results of remote sensing images, and then affect the width of the section. Moreover, water body identification based on remote sensing interpretation is more suitable for rivers with regular boundaries or wide sea-entering rivers and mountainous rivers. The identification error is large for narrow, winding and continuous rivers. Therefore, the error of the river section width extracted according to the remote sensing image interpretation results is large, and it cannot fully and accurately describe the river width. The correction of the section width still needs to be further used by visual interpretation or hydrological and topographic data query methods, which requires a lot of manpower. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings and defects of the existing river section width correction technology, such as over-reliance on flood measured data and errors in remote sensing image interpretation, and to provide an automatic correction method for river section width based on BG (Bernaola-Galvan) segmentation and MK (Maan-Kendall) trend test. The automatic correction of river section width is realized completely based on digital elevation model data, and can provide certain technical support for the river section processing process of hydrological and hydrodynamic numerical model construction, river engineering design and other operations.

[0005] A method for automatically correcting the width of a river section based on BG segmentation and MK trend test comprises the following steps:

[0006] S1. Calculation of slope and aspect in the study area;

[0007] S2. Generate initial river section;

[0008] S3. Extraction of cross-section information;

[0009] S4. MK trend test of section slope;

[0010] S5. BG mutation test of section slope;

[0011] S6. Adjust the section width according to the results of the slope MK trend test and the slope aspect BG mutation test.

[0012] Among them, in the step S1, the slope and aspect calculation of the study area includes calculating the slope and aspect of the study area based on the DEM data, using the fitting surface method, using a 3x3 window to traverse the entire DEM data, and for each central grid unit, calculating the horizontal distance d between the 8 adjacent grids of the current grid x and elevation difference d y , then based on the horizontal distance d x and elevation difference d y The slope and aspect are calculated, and the slope is divided into eight aspects.

[0013] Among them, the horizontal distance d of the grid e x and elevation difference d y The calculation is as follows:

[0014]

[0015] Wherein, in the step S1, the slope is obtained by calculating the ratio of the elevation difference and the horizontal distance between each grid cell and its adjacent grid cells;

[0016]

[0017] In the formula, S P is the slope percentage;

[0018] The aspect is obtained by calculating the direction of maximum rate of elevation change for each grid cell:

[0019]

[0020] In the formula, A s is the slope direction;

[0021] The eight slope directions include due north (0-22.5)(337.5-360), due south (157-202.5), due west (247.5-292.5), due east (67.5-112.5), northeast (22.5-67.5), southeast (112.5-157.5), southwest (202.5-247.5) and northwest (292.5-337.5).

[0022] Among them, in the step S2, the river centerline is drawn based on the water body identification result based on remote sensing interpretation, the river centerline is segmented according to the section interval requirements, and a river section perpendicular to the river centerline is generated at the breakpoint. The section width is consistent with the water body boundary, and the section shapefile data is obtained.

[0023] Wherein, in step S3, the section information extraction includes: reading the section data with the help of the open source raster spatial data conversion library (Geospatial Data Abstraction Library, GDAL) and superimposing each section line with the Dem, aspect and slope raster data, respectively extracting the elevation, aspect and slope profile lattice of each raster pixel on the section;

[0024] Wherein, in the step S4, the section is divided into left slope and right slope according to the lowest point of the slope direction / river channel section, and the slopes of the left slope and the right slope of each section are subjected to MK trend analysis to determine whether the section passes the MK trend test. If so, the shapefile file and the elevation lattice result of the section are output, if not, the process proceeds to step S5;

[0025] Wherein, in the step S4, the section is divided into a left slope and a right slope according to the slope direction / the lowest point of the river section, and the slopes of the left slope and the right slope of each section are subjected to MK trend analysis, including:

[0026] For the distance sequence x of the left or right slope i , construct the order sequence S of the slope k :

[0027]

[0028] r i Represents x i >x j The cumulative number of samples (1≤j≤i), assuming that the distance sequence is random and independent, calculates the standard normal distribution statistical variable UF k :

[0029]

[0030]

[0031] In the formula, E(S k ) is the slope order sequence S k The mean of For S k The variance of |UF k |≥U α , indicating that the slope has a preset trend change, U αIt indicates the critical value used to judge whether the trend change of time series is significant under the given significance level α; the distance sequence is arranged in reverse order and UB is calculated k sequence:

[0032] UB k =-UF k (k=n,n-1,…,1);

[0033] Judging UF k and UB k Whether there is an intersection in the sequence;

[0034] If the slope of the left and right slopes is UF k and UB k There is only one intersection point in each sequence. The coordinate before the left bank intersection and the coordinate after the right bank intersection are taken as the left and right endpoints of the river section by default. The section width is reduced and adjusted, and the shapefile file and elevation point matrix result of the section are output. If there are multiple intersection points or no intersection points, execute step 5.

[0035] Wherein, in step S5, the cross-section slope BG mutation test includes:

[0036] The slope directions of the left and right slopes of the river section that fails the MK trend test are obtained. If there are no pixels with opposite and continuous directions for the slope directions of the left and right slopes, the section is moved within a predetermined length range upstream and downstream along the river flow direction with the resolution of the grid data as the step size, and the length of the section is consistent with the water body range. Then, step S4 is repeated to extract the elevation, slope, and slope direction lattice of the moved section, and the MK trend test is performed on the left and right slopes again until there are pixels with opposite and continuous directions for the left and right slopes. Then, the slope direction lattice of the left and right slopes of each section is segmented by BG until it cannot be segmented.

[0037] The step of performing BG segmentation on the slope aspect lattices of the left slope and the right slope of each section includes:

[0038] Calculate the average slope U of the left and right slopes at the distance before and after each point i on the section L , U R , and the standard deviation of the slope S of the front and rear distances L , S R ,i=2,3,…,n-1;Calculate the average value of the left slope U k and the average right slope aspect U R The difference between them is T(i);

[0039]

[0040]

[0041] In the formula, S D is the combined deviation, N L and N R is the number of samples before and after point i. The larger T(i) is, the greater the difference in slope direction before and after point i is. The maximum value T in T(i) is calculated. m and statistical significance P(T m ):

[0042]

[0043] In the formula, γ=4.19lnN-11.54, δ=0.4, v=N-2, N is the distance length of the left slope or the right slope, is the square of the maximum value of the test statistic in time series analysis, I is the function operator, and the Beta distribution with parameters δv and δ is calculated. The complementary cumulative distribution function value at the point; set a critical value P0, the confidence level is 0.95, if the significance of a point P(T m )≥P k , then the distance sequence is divided into two segments at this point, otherwise it is not divided; the two slope subsequences obtained by segmentation are repeatedly calculated, and the statistical significance P(T m ) and judge until all subsequences can no longer be split.

[0044] Wherein, in step S6, adjusting the section width according to the slope MK trend test and the slope direction BG mutation test results includes:

[0045] If the slope of the left or right slope is UF k and UB k If there is no intersection in the sequence and there is one or more mutation points in the slope BG segmentation result, the length of the left slope or the right slope is shortened; the left slope or the right slope is shortened to the position of the next river section mileage point farthest from the slope BG segmentation mutation point. The default section width adjustment is completed, and the adjusted section shapefile and elevation dot matrix are output;

[0046] Among them, when shortening the left or right slope to the next river section mileage point of the farthest slope direction BG split mutation point, if the mileage of the left slope is [0, 11, 24, 32, 43, 55], where 11, 32, 43 are slope direction mutation points, shortening the left slope to the next river section mileage point of the farthest slope direction BG split mutation point means shortening the left slope to 55. Similarly, if the mileage of the right slope is [156, 172, 188, 197, 202, 223], where 188, 197, 202 are slope direction mutation points, shortening the right slope to the next river section mileage point of the farthest slope direction BG split mutation point means shortening the left slope to 172;

[0047] If the slope of the left or right slope is UF k and UB k If there is no intersection in the sequence and there is no mutation point in the slope BG segmentation result, the elevation, slope direction and slope of the section are re-extracted after extending the length of the left slope or right slope by one pixel, and the slope of the left slope or right slope of the adjusted section is subjected to MK trend analysis. When the length of the extended left slope or right slope is greater than the width of the section itself, the section is re-moved within a predetermined length range upstream and downstream along the river flow direction with the resolution of the grid data as the step size and consistent with the section length and water body range, and then step S4 is repeated to extract the elevation, slope and slope direction lattice of the moved section, and the left and right slopes are re-tested for MK trend until there are opposite and continuous pixels in the left and right slope directions, and then the slope direction lattice of the left and right slopes of each section is subjected to BG segmentation until it cannot be segmented;

[0048] If the UF of the left or right slope k and UB k If there are multiple intersections in the sequence, and the slope BG segmentation result of the left or right slope has one or more mutation points, the length of the left or right slope is shortened; the left or right slope is shortened to the position of the nearest slope BG segmentation mutation point, the elevation, slope direction and slope of the section are re-extracted, and the slope of the left or right slope of the adjusted section is subjected to MK trend analysis;

[0049] Among them, when shortening the left or right slope to the nearest slope direction BG split mutation point, if the mileage of the left slope is [0, 11, 24, 32, 43, 55], where 11, 32, 43 are slope direction mutation points, shortening the left slope to the nearest slope direction BG split mutation point means shortening the left slope to 11. Similarly, if the mileage of the right slope is [156, 172, 188, 197, 202, 223], where 188, 197, 202 are slope direction mutation points, shortening the right slope to the nearest slope direction BG split mutation point means shortening the left slope to 202; when UF k and UB k When there is only one intersection in the sequence, the section width is considered to be adjusted, and the adjusted section shapefile and elevation point matrix are output.

[0050] In a large number of cross-section tests, no UF on the left or right slope was found. k and UB k There are multiple intersections in the sequence, and there is no mutation point in the BG segmentation result of the left or right slope.

[0051] Since the change of river channel slope is periodic and nonlinear, the present invention uses Bernaola-Galvan mutation detection to more accurately capture the existence of mutation points, and the slope has a trend. The present invention uses Maan-Kendall test to effectively detect the trend characteristics of the slope.

[0052] The present invention adaptively generates sections of different widths according to water body identification results based on remote sensing interpretation, obtains mutation information of the section by comprehensively analyzing the non-stationarity of the slope direction and the monotonic trend of the slope, and then corrects the width of the river section.

[0053] The present invention comprehensively considers the periodic changes in the slope direction and the trend changes in the slope of the river channel, can more accurately analyze the morphological characteristics of the river channel section, and realizes automatic correction of the river channel section width. The method of the present invention has the advantages of being objective, easy to implement, and widely applicable to various regions. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a flow chart of a method for automatically correcting a river section width based on BG segmentation and MK trend test according to an embodiment of the present invention;

[0055] Figure 2 is a 3x3 window image used in an embodiment of the present invention to calculate the horizontal distance and elevation difference of any grid;

[0056] Figure 3 The river centerline and initial river section generated by the embodiment of the present invention according to the water body range;

[0057] Figure 4-1 It is an initial cross-sectional view of an initial river channel extracted from DEM data in an embodiment of the present invention (the horizontal axis represents the mileage of the river channel section, and the vertical axis represents the elevation of the river channel section);

[0058] Figure 4-2 MK trend test result diagram of the left slope of the initial river section of the embodiment of the present invention (the horizontal axis is the mileage of the river section);

[0059] Figure 4-3 It is a MK trend test result diagram of the right slope of the initial river section according to an embodiment of the present invention (the horizontal axis is the mileage of the river section); wherein the right slope passes the MK trend test;

[0060] Figure 5 It is an adjustment result diagram of the right slope of the section according to the slope MK trend test of the embodiment of the present invention;

[0061] Figure 6-1 It is a right slope MK trend test result diagram of the cross section of the embodiment of the present invention (the horizontal axis is the mileage of the river section);

[0062] Figure 6-2It is a BG segmentation result diagram of the slope aspect of the cross section of the embodiment of the present invention (the horizontal axis is the mileage of the river cross section, the vertical axis is the test statistic value T, and the vertical height represents the size of the statistic value T);

[0063] Figure 7-1 It is a river cross-section diagram after adjustment of the slope MK trend test and the slope BG segmentation result in the embodiment of the present invention (the horizontal axis represents the river cross-section mileage, and the vertical axis represents the river cross-section elevation);

[0064] Figure 7-2 , Figure 7-3 They are MK trend test result diagrams of the left slope and the right slope of the river section after adjustment in the embodiment of the present invention. DETAILED DESCRIPTION

[0065] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0066] The automatic correction method for river section width based on BG segmentation and MK trend test of the embodiment of the present invention includes the following implementation steps:

[0067] S1. Calculation of slope and aspect in the study area;

[0068] The slope and aspect of the study area were calculated based on DEM (digital elevation model) data, and the aspects were classified to obtain slope and aspect raster data respectively; the slope was divided into eight aspects.

[0069] Among them, the eight slope directions include due north (0-22.5)(337.5-360), due south (157-202.5), due west (247.5-292.5), due east (67.5-112.5), northeast (22.5-67.5), southeast (112.5-157.5), southwest (202.5-247.5) and northwest (292.5-337.5);

[0070] Specifically, the slope and aspect raster data are calculated based on the digital elevation model DEM data: the slope and aspect of the study area based on the DEM data are calculated using the fitting surface method, and the DEM data of the study area is traversed using a 3x3 window; Figure 2As shown, the grid cells of each 3x3 range of the 3x3 window are composed of the central grid cell with pixel value e in the middle of the square of equal length arranged in the positive direction, and the left grid cell with pixel value e1, the grid cell directly above with pixel value e2, the grid cell on the right with pixel value e3, the grid cell directly below with pixel value e4, the left oblique upper grid cell with pixel value e5, the right oblique upper grid cell with pixel value e6, the right oblique lower grid cell with pixel value e7, and the left oblique lower grid cell with pixel value e8. The grid is formed by the overlapping connection of the sides of each adjacent grid cell; the horizontal distance d between the central grid cell of each DEM data and the left grid cell, the grid cell directly above, the grid cell on the right, the grid cell directly below, the left oblique upper grid cell, the right oblique upper grid cell, the right oblique lower grid cell, and the left oblique lower grid cell is calculated by sliding the 3x3 window. x and elevation difference d y , generate slope and aspect raster data for subsequent river cross-section information extraction.

[0071] S2. Generate initial river section;

[0072] Draw the river centerline based on the water body identification results based on remote sensing interpretation, divide the river centerline into sections according to the section interval requirements, generate river sections perpendicular to the centerline at the breakpoints, and the positions of the left and right banks of the section are consistent with the water body boundary to obtain the section shapefile data, such as Figure 3 As shown; according to the water body range, as shown Figure 3 The length shown is the centerline of the river in the study area; according to the project's requirement for 1000-meter intervals between river cross sections in the study area, the centerline of the river in the study area is segmented to obtain multiple centerline breakpoints of the river in the study area, and at the multiple centerline breakpoints of the river in the study area, cross sections of the rivers in the study area that are perpendicular to the centerline of the river sections where the centerline breakpoints of the river in the study area are located are made respectively. The width of each river cross section in the study area is flush with the water body boundary of its river mileage, and shapefile data of each river cross section in the study area are obtained.

[0073] S3. Extraction of cross-section information;

[0074] Traverse and read the cross-section data and overlay each cross-section line with the Dem, aspect and slope grid data, and extract the elevation, aspect and slope profile lattice of each grid pixel on the cross section;

[0075] Specifically, the open source raster spatial data conversion library GDAL was used to read the shapefile data of river cross sections in multiple study areas, and then each river cross section line in multiple study areas was superimposed with the DEM data, aspect and slope raster data of the study areas, and the elevation, aspect and slope profile point matrix of each raster pixel on the river cross sections in multiple study areas were extracted.

[0076] S4. MK trend test of section slope;

[0077] Each section is divided into left slope and right slope with the pixel with slope aspect as plane (slope aspect value is -1) as the dividing line. If there is no slope aspect value of -1, the pixel with the lowest elevation of the section is used as the dividing line. The MK trend analysis is performed on the slopes of the left and right slopes of each section.

[0078] Calculate the standard normal distribution statistic UF of the slope profile points of the cross-sections of the river channels in multiple study areas k and the reverse order statistic UB k Sequence; divide the cross sections of the river channels in multiple study areas into left slope and right slope according to the slope direction; calculate the standard normal distribution statistic UF of the left slope and right slope profile lattice of the cross sections of the river channels in multiple study areas respectively k and the reverse order statistic UB k sequence;

[0079] Judging UF k and UB k Does the sequence have an intersection? If the slopes of the left and right slopes are UF k and UB k There is only one intersection point in each sequence. The coordinate before the left bank intersection and the coordinate after the right bank intersection are taken as the left and right endpoints of the river section by default, and the section width is reduced and adjusted. Figure 4-1 , Figure 4-2 as well as Figure 4-3 As shown in the figure, the right slope of the section passed the consistency test of MK. After the adjustment, the right slope Figure 5 As shown; output the shapefile files and elevation point matrix results of each river cross section in the study area that passed the MK trend test.

[0080] S5. BG mutation test of section slope;

[0081] For sections that fail the MK trend test, first determine the slope directions of the left and right slopes based on the flow direction of the river section. If there are no pixels with opposite and continuous directions on the left and right slopes, move the section within 100m upstream and downstream along the river flow direction with the resolution of the grid data as the step size, and keep the section length consistent with the water body range. Then extract the elevation and slope lattices of the moved section and re-perform the MK trend test on the left and right slopes until there are opposite and continuous pixels on the left and right slopes. If there are continuous pixels in the left and right slopes, perform BG segmentation on the slope lattices of the left and right slopes of each section.

[0082] S6. Adjust the section width according to the results of the slope MK trend test and the slope aspect BG mutation test.

[0083] like Figure 6-1 , Figure 6-2 As shown, in this application, according to the MK trend analysis results and BG segmentation results of the left slope of the section, the UF k and UB k If there are multiple intersections in the sequence, it is necessary to shorten the cross-section length of the left slope, shorten the left bank endpoint to the position of the nearest slope BG segmentation mutation point, and re-perform the slope MK trend test on the left slope. Based on the test results, it is determined whether to perform the slope BG segmentation. Figure 7-1 , Figure 7-2 as well as Figure 7-3 Shown is the cross-section result after the final shortening of the left slope.

[0084] Depend on Figure 7-1 , Figure 7-2 as well as Figure 7-3 It can be seen that the standard normal distribution statistical variable UF of the left slope profile lattice of a river cross section is k and the reverse order statistic UB k The standard normal distribution statistical variable UF of the sequence and right slope profile lattice k and the reverse order statistic UB k There is only one intersection point in each sequence, and the cross-sectional width of the river is adjusted by default.

[0085] Comparing the river cross-section before and after adjustment, it can be seen that the width correction effect of the left and right banks of the river cross-section is good.

[0086] The embodiments of the present invention illustrate that the automatic correction method for river cross-sectional width of BG segmentation and MK trend test of the present invention is completely based on digital elevation model data, and combines the MK trend test method and the BG segmentation method to correct the river cross-sectional width.

[0087] Compared with other methods of river width extraction and correction, this method eliminates the process of searching for river hydrological data and identifies the changing trend and abnormal points of river section elevation based entirely on the trend of slope and the periodicity and nonlinearity of slope direction. The purpose of fast calculation is achieved by directly calculating with DEM data.

[0088] Therefore, the automatic correction method of the river cross-sectional width of the BG segmentation and MK trend test of the present invention has the characteristics of being independent of historical hydrological data, rich in physical meaning, fast in calculation, and applicable to a wide range of areas.

[0089] The basic principles and main features of the present invention and the advantages of the present invention are shown and described above. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention.

[0090] Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the claims are included in the present invention.

[0091] In addition, it should be understood that although the present 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 embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. The automatic correction method of river section width based on BG segmentation and MK trend test is characterized by: Includes steps: S1. Calculation of slope and aspect in the study area; S2. Generate initial river section; S3. Extraction of cross-section information; S4. MK trend test of section slope; S5. BG mutation test of section slope; S6. Adjust the section width according to the results of the slope MK trend test and the slope aspect BG mutation test.

2. The automatic correction method for river section width based on BG segmentation and MK trend test according to claim 1 is characterized in that: In step S1, the slope and aspect calculation of the study area includes: calculating the slope and aspect of the study area based on DEM data, using the fitting surface method, using a 3x3 window to traverse the entire DEM data, and for each central grid unit, calculating the horizontal distance d between the eight adjacent grids of the current grid. x and elevation difference d y , then based on the horizontal distance d x and elevation difference d y The slope and aspect are calculated, and the slope is divided into eight aspects.

3. The automatic correction method for river section width based on BG segmentation and MK trend test according to claim 2 is characterized in that: In step S1, the horizontal distance d of the grid e x and elevation difference d y The calculation is as follows: The slope is obtained by calculating the ratio of the elevation difference and the horizontal distance between each grid cell and its adjacent grid cells; In the formula, S P is the slope percentage; The slope aspect is obtained by calculating the maximum direction of the elevation change rate of each grid cell; In the formula, A s is the slope direction; The eight slope directions include due north (0-22.5)(337.5-360), due south (157-202.5), due west (247.5-292.5), due east (67.5-112.5), northeast (22.5-67.5), southeast (112.5-157.5), southwest (202.5-247.5) and northwest (292.5-337.5).

4. The automatic correction method for river section width based on BG segmentation and MK trend test according to claim 3 is characterized in that: In step S2, the river centerline is drawn based on the water body identification result based on remote sensing interpretation, and the river centerline is segmented according to the section interval requirement. A river section perpendicular to the river centerline is generated at the breakpoint, and the section width is consistent with the water body boundary to obtain section shapefile data.

5. The automatic correction method for river section width based on BG segmentation and MK trend test according to claim 4 is characterized in that: In step S3, the section information extraction includes: reading the section data with the help of an open source raster spatial data conversion library and superimposing each section line with the Dem, aspect and slope raster data, and respectively extracting the elevation, aspect and slope profile lattice of each raster pixel on the section.

6. The method for automatically correcting the width of a river section based on BG segmentation and MK trend test according to claim 5 is characterized in that: In step S4, the MK trend test of the section slope includes: dividing the section into a left slope and a right slope according to the slope direction / the lowest point of the river section, performing an MK trend analysis on the slopes of the left slope and the right slope of each section, and judging whether the section passes the MK trend test. If so, outputting the shapefile file and the elevation dot matrix result of the section, if not, entering step S5.

7. The automatic correction method for river section width based on BG segmentation and MK trend test according to claim 6 is characterized in that: In step S4, the section is divided into a left slope and a right slope according to the slope direction / the lowest point of the river section, and the slopes of the left slope and the right slope of each section are subjected to MK trend analysis, including: For the distance sequence x of the left or right slope i , construct the order sequence S of the slope k : r i Represents x i >x j The cumulative number of samples (1≤j≤i), assuming that the distance sequence is random and independent, calculates the standard normal distribution statistical variable UF k : In the formula, E(S k ) is the slope order sequence S k The mean of For S k The variance of |UF k |≥U α , indicating that the slope has a preset trend change, U α It indicates the critical value used to judge whether the trend change of time series is significant under the given significance level α; the distance sequence is arranged in reverse order and UB is calculated k sequence: UB k =-UF k (k=n,n-1,…,1); Determine UB k and UB k Whether there is an intersection in the sequence; If the slope of the left and right slopes is UF k and UB k There is only one intersection point in each sequence. The coordinate before the left bank intersection and the coordinate after the right bank intersection are taken as the left and right endpoints of the river section by default. The section width is reduced and adjusted, and the shapefile file and elevation point matrix result of the section are output. If there are multiple intersection points or no intersection points, execute step 5.

8. The automatic correction method for river section width based on BG segmentation and MK trend test according to claim 7 is characterized in that: In step S5, the cross-section slope BG mutation test includes: The left and right slope directions of the river section that fails the MK test are obtained. If there are no pixels with opposite and continuous directions on the left and right slopes, the section is moved within a predetermined length range upstream and downstream along the river flow direction with the resolution of the grid data as the step size, and the length of the section is consistent with the water body range. Then step S4 is repeated to extract the elevation, slope, and slope direction lattice of the moved section, and the MK trend test is performed on the left and right slopes again until there are pixels with opposite and continuous directions on the left and right slopes. Then, the slope direction lattice of the left and right slopes of each section is segmented by BG until it cannot be segmented.

9. The automatic correction method for river section width based on BG segmentation and MK trend test according to claim 9 is characterized in that: The BG segmentation of the slope aspect lattice of the left slope and the right slope of each section includes: Calculate the average slope U of the left and right slopes at the distance before and after each point i on the section L , U R , and the standard deviation of the slope S of the front and rear distances L , S R ,i=2,3,…,n-1;Calculate the average value of the left slope U L and the average right slope aspect U R The difference between them is T(i); In the formula, S D is the combined deviation, N L and N R is the number of samples before and after point i. The larger T(i) is, the greater the difference in slope direction before and after point i is. The maximum value T in T(i) is calculated. m and statistical significance P(T m ): Where, γ=4.19lnN-11.54, δ=0.4, v=N-2, N is the distance length of the left slope or right slope; T max2 is the square of the maximum value of the test statistic in time series analysis, I is the function operator, and the Beta distribution with parameters δv and δ is calculated. The complementary cumulative distribution function value at the point; set a critical value P0, the confidence level is 0.95, if the significance of a point P(T m )≥P0, the distance sequence is divided into two segments at this point, otherwise it is not divided; the two slope subsequences obtained by segmentation are repeatedly calculated, and the statistical significance P(T m ) and judge until all subsequences can no longer be split.

10. The method for automatically correcting the width of a river section based on BG segmentation and MK trend test according to claim 9, characterized in that: In step S6, adjusting the section width according to the slope MK trend test and the slope aspect BG mutation test results includes: If the slope of the left or right slope is UF k and UB k If there is no intersection in the sequence and there is one or more mutation points in the slope BG segmentation result, the length of the left slope or the right slope is shortened; the left slope or the right slope is shortened to the position of the next river section mileage point farthest from the slope BG segmentation mutation point. The default section width adjustment is completed, and the adjusted section shapefile and elevation dot matrix are output; If the slope of the left or right slope is UF k and UB k If there is no intersection in the sequence and there is no mutation point in the slope BG segmentation result, the elevation, slope direction and slope of the section are re-extracted after extending the length of the left slope or the right slope by one pixel, and the slope of the left slope or the right slope of the adjusted section is subjected to MK trend analysis; when the length of the extended left slope or the right slope is greater than the width of the section itself, the section is re-moved within a predetermined length range upstream and downstream along the river flow direction with the resolution of the grid data as the step length and consistent with the section length and the water body range, and then step S4 is repeated to extract the elevation, slope and slope direction lattice of the moved section, and the left and right slopes are re-tested for MK trend until there are opposite and continuous pixels in the left and right slope directions, and then the slope direction lattice of the left and right slopes of each section is subjected to BG segmentation until it cannot be segmented; If the UF of the left or right slope k and UB k If there are multiple intersections in the sequence, and the slope BG segmentation result of the left or right slope has one or more mutation points, the length of the left or right slope is shortened; the left or right slope is shortened to the position of the nearest slope BG segmentation mutation point, the elevation, slope direction and slope of the section are re-extracted, and the slope of the left or right slope of the adjusted section is subjected to MK trend analysis; When UF k and UB k When there is only one intersection in the sequence, the section width is considered to be adjusted, and the adjusted section shapefile and elevation point matrix are output.

Images