Bedding landslide hidden danger identification method based on slope body structure and multi-source remote sensing technology

Abstract

The invention discloses a bedding landslide hidden danger identification method based on a slope body structure and a multi-source remote sensing technology, and relates to the technical field of engineering geological slope disaster prevention and control. The method comprises the following steps: acquiring an elevation image and historical landslide data of a research area, and carrying out unit division on the research area to obtain each slope unit in the research area; screening is conducted according to the rock stratum tendency, the slope included angle and the number of free faces; disaster control factors of landslide disasters of the bedding slope are extracted; determining an information amount value of each disaster control factor based on an information amount model by taking the disaster control factors as input parameters of the information amount model; all the information amount values are superposed based on a grid calculator, and the susceptibility index of each grid is determined; obtaining deformation data from partitions with high susceptibility in the susceptibility partition map; and determining the work point with the bedding landslide hidden danger in the research area by combining the optical remote sensing image and the susceptibility partition map. The method can improve the recognition efficiency and accuracy of the bedding landslide hidden danger.

Description

Technical Field

[0001] The present application relates to the technical field of engineering geological slope disaster prevention and control, and specifically to a method for identifying the hidden dangers of layered landslides based on slope structure and multi-source remote sensing technology. Background Art

[0002] With the rapid development of infrastructure construction, a large number of rock slope projects have emerged in the fields of transportation, mining, hydropower and other engineering fields, among which the problem of bedding landslides has become increasingly prominent. Bedding landslide refers to the sliding of the slope along the surface whose inclination direction is close to or roughly consistent with the inclination of the layered bedrock. Bedding landslides are characterized by easy instability, large scale and difficult to prevent. In particular, weak interlayers such as interlayer dislocation zones and mud layers are often developed in the rock mass of red-bed slopes, which is a typical "easy-to-slide bedding structural surface". Frequent bedding landslide disasters pose a serious threat to people's lives and property, engineering construction safety and the operation of important infrastructure. Research on the identification and evaluation method of bedding landslide hazards is of great significance to disaster prevention and mitigation work in this region.

[0003] At present, the engineering community generally believes that bedding landslide refers to the bedding sliding damage that occurs on the layered structure slope where the angle between the rock layer and the slope direction is less than 30 degrees and the slope is greater than the rock layer inclination angle. However, in highway construction, it is found that in areas with developed surface water systems, the front edge or side wall of the bedding slope is occasionally eroded to form an open surface, and new open positions are formed after the slope is excavated, that is, there is an unfavorable spatial structure with double or even three sides facing open. Traditional identification of landslide hazards mostly relies on on-site investigations by professional technicians, and comprehensive judgments are made through surface cracks or subsidence, slope sliding deformation, surface vegetation dumping, groundwater seepage characteristics, etc., and it relies heavily on engineering experience. In recent years, with the continuous development of computer information technology, advanced methods such as geographic information system (GIS), remote sensing interpretation (RS), and interferometric synthetic aperture radar (InSar) have been gradually applied to the identification and evaluation of landslide hazards, and relatively systematic research methods and results have been achieved. However, a single method is limited by complex terrain and geological conditions in rugged mountainous areas. For example, when the vegetation coverage rate is high, it is difficult for image data to accurately obtain surface information, and the landslide susceptibility assessment method also focuses on the comprehensive topography, stratigraphic rock groups, etc., without considering the slope structure of the bedding landslide, that is, the combined relationship between the rock layer occurrence and the geometric characteristics of the slope. Therefore, it is not significantly applicable to the identification and evaluation of bedding landslides in red soft rocks. Summary of the invention

[0004] The present application provides a method for identifying landslide hazards along the layers based on slope structure and multi-source remote sensing technology. By integrating multi-source remote sensing data with dynamic analysis of slope structure, accurate identification and graded early warning of landslide hazards along the layers can be achieved, thereby improving the efficiency and accuracy of identifying landslide hazards along the layers.

[0005] The method for identifying landslide hazards along the layer based on slope structure and multi-source remote sensing technology provided in the embodiment of the present application may include: Obtain elevation images and historical landslide data of the study area to divide the study area into units, and obtain each slope unit in the study area; In the slope unit, screening is performed according to the stratum inclination, the angle of slope direction and the number of free faces to obtain the bedding slope; Extracting the controlling factors of the landslide disaster of the bedding slope; Taking the disaster control factor as an input parameter of an information volume model, determining the information volume value of each of the disaster control factors based on the information volume model; Based on the grid calculator, all the information values ​​are superimposed to determine the susceptibility index of each grid, and the susceptibility index is divided into a plurality of classification intervals to obtain a susceptibility zoning map of the bedding landslide in the study area; Acquire deformation data from a partition with high susceptibility in the susceptibility partition map, and verify the partition result of the susceptibility partition based on the deformation data; The optical remote sensing images and the susceptibility zoning map are combined to determine the construction sites in the study area that have the risk of bedding landslide.

[0006] In some embodiments, the step of obtaining the elevation image and historical landslide data of the study area to divide the study area into units to obtain each slope unit in the study area may include: Acquire an elevation image and historical landslide data of a study area, and extract a slope aspect and a mountain shadow map from the elevation image and the historical landslide data respectively; Determining parameters of a multi-scale segmentation algorithm based on the slope aspect and the hill shadow map; the parameters of the multi-scale segmentation algorithm include scale, shape feature weight and compactness weight parameters; Combining the modified trial and error method with the parameters of the multi-scale segmentation algorithm, each slope unit in the study area is obtained after all areas are merged through pixel aggregation and heterogeneity threshold merging.

[0007] In some embodiments, in the slope unit, screening is performed according to the stratum inclination, the angle of the slope direction and the number of free faces to obtain the bedding slope, including: Slope units whose angle between rock stratum dip and slope aspect is less than the set threshold are classified as forward slopes, and the threshold range of the angle is adjusted according to the conditions of single-sided empty space, double-sided empty space, and three-sided empty space.

[0008] In some embodiments, the slope unit whose angle between the stratum dip and the slope direction is less than a set threshold is classified as a forward slope, and the threshold range of the angle is adjusted according to the conditions of single-sided empty space, double-sided empty space, and three-sided empty space, including: The slope structure combinations with dip angles less than 10 degrees are classified as nearly horizontal structure slopes; Generate the interpolated rock stratum dip by Kriging interpolation, and perform raster calculation on the interpolated rock stratum dip and the slope aspect to obtain the included angle between the rock stratum dip and the slope aspect; For the forward slope with one side facing the sky, retain the slope construction points with an included angle less than 30 degrees; for the forward slope with two sides facing the sky, retain the slope construction points with an included angle less than 60 degrees; for the forward slope with three sides facing the sky, retain the slope construction points with an included angle less than 90 degrees.

[0009] In some embodiments, the extraction of the disaster control factors for the landslide disasters of the bedding slopes includes: Calculate the frequency ratios of the topographic and geomorphic features, stratigraphic lithology, geological structure, hydro-meteorology, and engineering disturbance indicators within the bedding slopes, and screen the 5 factors with the largest frequency ratios as the disaster control factors.

[0010] In some embodiments, taking the disaster control factors as the input parameters of the information model, and determining the information values of each disaster control factor based on the information model includes: Let each disaster control factor be X i ( i = 1, 2, 3...), for the landslide events H The information values provided I ( X i , H ) can be calculated in the form of sample frequencies:

[0011] Among them, N is the total number of landslide disasters in the study area; T is the total number of evaluation units in the study area; is the number of landslides occurring in the i th category of the j rd disaster control factor; is the number of evaluation units occupied by the i th category in the j th disaster control factor.

[0012] In some embodiments, based on the raster calculator to superimpose all the information values, determine the susceptibility index of each raster, divide the susceptibility index into multiple grading intervals, and obtain the susceptibility zoning map of the bedding landslides in the study area, including: Based on the ArcGIS raster calculator, superimpose the information values of each disaster control factor to obtain the susceptibility index of each raster.

[0013] In some embodiments, the grid-based calculator superimposes all the information values, determines the susceptibility index of each grid, divides the susceptibility index into a plurality of grading intervals, and obtains a susceptibility zoning map of the bedding landslide in the study area, further comprising: The natural breakpoint method was used to divide the susceptibility of bedding landslides in the study area into four classification intervals, and a susceptibility zoning map of bedding landslides in the study area was obtained.

[0014] In some embodiments, the obtaining of deformation data from the partition with high susceptibility in the susceptibility partition map, and verifying the partition result of the susceptibility partition based on the deformation data, includes: InSAR deformation monitoring is performed on the bedding slope in the high-susceptibility zone, and deformation time series data is obtained from the high-susceptibility zone in the susceptibility zone map.

[0015] In some embodiments, the InSAR deformation monitoring includes: The interference pairs of the slope along the bedding in the high-prone area are differentially processed, and the nonlinear deformation and noise phase are separated by spatiotemporal filtering. The deformation time series is reconstructed by combining the weighted least squares method and the temporal coherence coefficient, and the temporal coherence coefficient is used to screen the coherent points.

[0016] Compared with the existing technology, the beneficial effects of the present application are: by integrating multi-source remote sensing data and dynamic analysis of slope structure, accurate identification and graded early warning of bedding landslide hazards can be achieved, bedding slopes can be quickly locked by screening based on terrain segmentation and rock formation occurrence, spatial differences in susceptibility can be quantified by combining the disaster control factor information model, and a closed-loop verification mechanism is formed by deformation monitoring and optical image feature verification. This effectively solves the problems of traditional methods relying on manual experience, ignoring the particularity of slope structure, and insufficient recognition accuracy in complex mountainous environments, improves the recognition efficiency and accuracy of bedding landslide hazards, and provides data-driven technical support for engineering construction and disaster prevention and control in dangerous mountainous areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic diagram of the steps of a method for identifying landslide hazards along the layer based on slope structure and multi-source remote sensing technology provided in an embodiment of the present application.

[0018] Figure 2 This is a schematic diagram of the process of identifying landslide hazards along the layer based on slope structure and multi-source remote sensing technology in an embodiment of the present application.

[0019] Figure 3 This is an image of the study area provided in the examples of this application.

[0020] Figure 4 A schematic diagram of the slope unit division provided in an embodiment of the present application.

[0021] Figure 5 A slope structure zoning diagram provided for an embodiment of the present application.

[0022] Figure 6 A susceptibility zoning map of bedding landslides in the study area provided in the embodiments of the present application.

[0023] Figure 7 A schematic diagram of the InSAR deformation rate provided in an embodiment of the present application.

[0024] Figure 8 A schematic diagram of the final determination of a bedding landslide provided in an embodiment of the present application. DETAILED DESCRIPTION

[0025] The present application is further described in detail below in conjunction with test examples and specific implementation methods. However, this should not be understood as the scope of the above subject matter of the present application being limited to the following embodiments, and all technologies implemented based on the content of the present application belong to the scope of protection of the present application.

[0026] Unless otherwise specified, in the description of the specific embodiments of the present application, the terms indicating the orientation or position relationship such as "up", "down", "left", "right", "center", "inside", "outside", "side", etc. are all expressions based on the orientation or position relationship shown in the drawings, or the orientation or position relationship when the product / equipment / device is usually used. These terms of orientation or position relationship are only for the convenience of describing the scheme of the present application or simplifying the description in the specific embodiments to facilitate the technicians to quickly understand the scheme, rather than indicating or implying that a specific device / component / element must have a specific orientation, or be constructed and operated in a specific position relationship, and therefore cannot be understood as a limitation on the present application.

[0027] In the description of the embodiments of the present application, the technical terms "first", "second", etc. only distinguish one entity or operation from another entity or operation, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0028] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0029] Please see Figure 1 , Figure 1 A schematic diagram of the steps of a method for identifying hazard of landslide along the layer based on slope structure and multi-source remote sensing technology provided in an embodiment of the present application. The method for identifying hazard of landslide along the layer based on slope structure and multi-source remote sensing technology may include: S1. Obtain elevation images and historical landslide data of the study area to divide the study area into units and obtain each slope unit in the study area.

[0030] S2. In the slope unit, screening is performed according to the inclination of the rock strata, the angle of the slope direction and the number of free surfaces to obtain the bedding slope.

[0031] S3. Extract the controlling factors of landslide disasters in the layer slope.

[0032] S4. Taking the disaster control factor as the input parameter of the information quantity model, the information quantity value of each disaster control factor is determined based on the information quantity model.

[0033] S5. Based on the grid calculator, all information values ​​are superimposed to determine the susceptibility index of each grid, and the susceptibility index is divided into multiple classification intervals to obtain the susceptibility zoning map of bedding landslides in the study area.

[0034] S6. Obtain deformation data from the partition with high susceptibility in the susceptibility partition map, and verify the partition result of the susceptibility partition based on the deformation data.

[0035] S7. Combine optical remote sensing images and susceptibility zoning maps to determine the construction sites in the study area that have the risk of landslides in the bedding layer.

[0036] Please Figure 1 Based on the reference Figure 2 , Figure 2 This is a schematic diagram of the process of identifying landslide hazards along the layer based on slope structure and multi-source remote sensing technology in an embodiment of the present application.

[0037] The specific implementation of step S1 may include: Acquire an elevation image and historical landslide data of a study area, and extract a slope aspect and a mountain shadow map from the elevation image and the historical landslide data respectively; Determining parameters of a multi-scale segmentation algorithm based on the slope aspect and the hill shadow map; the parameters of the multi-scale segmentation algorithm include scale, shape feature weight and compactness weight parameters; Combining the modified trial and error method with the parameters of the multi-scale segmentation algorithm, each slope unit in the study area is obtained after all areas are merged through pixel aggregation and heterogeneity threshold merging.

[0038] First, we collected the basic elevation images and historical landslide database of the study area, extracted the slope direction reflecting the terrain trend and the hill shadow map enhancing the terrain undulation characteristics from the data elevation (DEM) and historical landslide database, and extracted the landslide shape and size characteristics as the basic input of the Multi-Scale Segmentation (MSS) method, and counted the morphology and scale characteristics of all historical landslides, including stratum lithology, waviness, slope direction, etc. Figure 3 , Figure 3 This is an image of the study area provided in the examples of this application.

[0039] Then, the slope units were divided by multi-scale segmentation method, and the parameters of the multi-scale segmentation algorithm, including scale, shape feature weight and compactness weight parameters, were set by trial and error method and combined with the morphological and scale characteristics of historical landslides in the study area (modified trial and error method). The scale parameter controls the minimum heterogeneity threshold of the segmentation unit, and the shape weight and compactness weight respectively constrain the geometric regularity of the segmented object.

[0040] On this basis, the algorithm starts from the pixel layer, and first aggregates pixels with similar features in adjacent regions into small image regions based on the modified trial and error method and the set scale, shape and compact weight parameters; similar small image regions are merged into large image regions based on the principle of minimum heterogeneity. For each merger, it is calculated whether the heterogeneity f of the merged region is greater than the scale threshold s; if it is greater than the scale threshold s, the two regions are not merged; if it is less than the scale threshold s, they are merged to generate a new larger image region; for example, the heterogeneity f of the merged region is calculated for the first merger. 1 Is it greater than the scale threshold s? Until the nth merging is performed based on the above results, the heterogeneity f of the merged region of the two regions is calculated after the merging. n Is it greater than the scale threshold? If so, perform the n+1th merge based on the above results. At this time, if the heterogeneity of all regions is greater than the scale threshold, or all regions have been merged, stop merging. Perform segmentation, merging, smoothing and other optimization processing on the image objects to eliminate the objects with local "island effect" and finally obtain the slope units in the study area. Please refer to Figure 4 , Figure 4 A schematic diagram of the slope unit division provided in an embodiment of the present application.

[0041] By evaluating the slope structure and susceptibility of hidden dangers of slope units, using InSAR technology to quantitatively identify landslide hazards based on spatiotemporal characteristics such as deformation rate, cumulative deformation magnitude, and deformation area boundaries, and finally combining optical remote sensing images to determine the work points in the study area with hidden dangers of bedding landslides, the bedding landslide identification results are obtained.

[0042] The specific implementation of step S2 may include: Slope units whose angle between rock stratum dip and slope aspect is less than the set threshold are classified as forward slopes, and the threshold range of the angle is adjusted according to the conditions of single-sided empty space, double-sided empty space, and three-sided empty space.

[0043] Specifically, slopes with an inclination angle of less than 10 degrees are classified as near-horizontal structural slopes, because the inclination angle of the rock layer is less than 10 degrees, close to the horizontal state. Such slopes have a low sliding risk due to the near-horizontal rock layer, and can be directly excluded from the analysis of bedding landslides.

[0044] The interpolated stratum dip is generated by Kriging interpolation, and the interpolated stratum dip and the slope direction are rasterized to obtain the angle between the stratum dip and the slope direction; for a single-sided open-air forward slope, slope work points with an angle less than 30 degrees are retained; for a double-sided open-air forward slope, slope work points with an angle less than 60 degrees are retained; for a three-sided open-air forward slope, slope work points with an angle less than 90 degrees are retained.

[0045] Among them, ArcGIS software can be used to estimate the occurrence between data points through Kriging interpolation method. When predicting the slope structure data of unknown points based on the rock formation occurrence and slope aspect, slope gradient and other slope structure data of existing slope data points, the covariance function of formula (1) and formula (2) is used:

[0046]

[0047] In the formula, is the Euclidean distance between data points i and j, represents the spatial correlation between data points i and j, ( , )and( , ) is the plane coordinate of the i-th slope data point and the j-th slope data point; Z( , ) and Z( , ) are the slope structure characteristic values ​​of the ith data point and the jth data point, such as rock layer inclination, slope aspect, slope, etc.

[0048] The spatial correlation of the slope structure characteristic values ​​at different locations is obtained by and The Gaussian function relationship between them is used to measure, as shown in formula (3): (3) Where: C 0, C and a are nugget constant, arch height and range respectively; h is the lag distance, which indicates the spatial distance between two points and is used to calculate the covariance; e is the base of the natural logarithm. Based on this, the covariance function relationship between the predicted point of the slope structure characteristic value and the existing numerical point is as follows: (4) Where: n is the total number of known data points, which is determined based on the data in the actual slope hazard identification application; represents the spatial correlation between data points i and j, This is the special case when i=j=n, usually reflecting the inherent variability of the data; is the weighted contribution of the i-th slope structure characteristic value to the slope structure characteristic value of the prediction point. According to the known point eigenvalue Z 1 ~Z n Calculation is as follows (5):

[0049] Among them, k is an index variable, indicating the kth data point, and the value range of k is k=1, 2,…, n; is the slope structure characteristic value of the kth known data point, It is the weight coefficient calculated in Kriging interpolation, which represents the kth known data point to the predicted point Contribution weight; for different slope structure characteristic value data of the prediction point, such as rock layer inclination, slope direction, slope gradient, etc., formula (1) to formula (5) are implemented respectively to obtain the slope structure characteristic value of the prediction point.

[0050] The interpolated rock layer inclination and slope direction are grid-calculated to obtain the angle β between the two. According to the β value, the slope structure is further divided into three types: forward slope (β<45 degrees), oblique slope (45 degrees<β<135 degrees), and reverse slope (β>135 degrees). For the selected forward slopes, according to the existing open-air conditions of the slopes and the excavation conditions of highway construction, the forward slopes are divided into three types: single-sided open-air (such as slope foot excavation), double-sided open-air (such as slope foot excavation and river cutting) and three-sided open-air (such as slope foot excavation, river cutting and structural erosion). For the forward slopes with single-sided open-air, only β Slope work point <30 degrees; for the forward slope with empty space on both sides, keep β Slope work point <60 degrees; for the forward slope with three sides facing the open, keep β All the down-slope working points with a degree less than 90; all the down-slopes screened in this step are classified as down-slopes. Obtain the slope structure zoning map of the study area. Please refer to Figure 5 , Figure 5A slope structure zoning diagram provided for an embodiment of the present application.

[0051] The specific implementation of step S3 may include: The frequency ratios of topography, stratum lithology, geological structure, hydro-meteorological and engineering disturbance indicators in the bedding slope are calculated, and the five factors with the largest frequency ratios are selected as disaster control factors.

[0052] For the screened bedding slopes, the frequency ratio method was used to analyze the action mechanism of landslide hazard-causing factors on the bedding slopes in the study area, taking topography (including surface undulation, elevation, slope, slope direction, etc.), stratum lithology (including stratum age, combination of soft and hard rock types, stratum occurrence, etc.), geological structure (distance from the tectonic belt), hydrometeorological conditions (distance from the surface water system, rainfall, rainfall duration, etc.), and engineering disturbance (distance from the highway) as indicators.

[0053] Specifically: Assume that a certain influencing factor of landslide is divided into i categories (for example, the elevation factor is divided into 0~1000m, 1000m~2000m, 2000m~3000m... a total of i categories), B i is the landslide area of ​​the i-th type of a certain influencing factor, A i is the research area of ​​the i-th category of a certain impact factor, B is the total landslide area in the study area, A is the total area of ​​the study area, then the frequency ratio of the influencing factor of the slope along the layer is FR i Frequency Ratio is defined as:

[0054] According to the frequency calculation results of each factor, the main hazard factors of landslide disasters in the bedding slope of the study area are obtained by comparison. Assuming that the hazard factors of landslide disasters in the bedding slope of a certain study area are ranked as follows: slope> rock formation> soft and hard rock combination> engineering disturbance> rainfall> ...> surface undulation, the five hazard factors with the largest frequency ratio are taken as the control factors to evaluate the susceptibility of landslide disasters in the bedding slope.

[0055] The specific implementation of step S4 may include: According to the obtained controlling factors of landslide disasters along the bedding slope, such as "slope, rock formation, soft and hard rock combination, engineering disturbance, and rainfall", the information volume model is used to evaluate the susceptibility of landslide disasters along the bedding slope. X i ( i =1,2,3……), for landslide events H The amount of information provided I ( Xi , H ) can be calculated using the sample frequency form:

[0056] in, N is the total number of landslide hazards in the study area; T is the total number of evaluation units in the study area; For the i The first of the three controlling factors j The number of landslides of the type; For the i Among the disaster-controlling factors j The number of evaluation units occupied by the class.

[0057] The specific implementation of step S5 may include: The information value of each disaster control factor is superimposed based on the ArcGIS raster calculator to obtain the susceptibility index of each raster.

[0058] The natural breakpoint method was used to divide the susceptibility of bedding landslides in the study area into four classification intervals, and a susceptibility zoning map of bedding landslides in the study area was obtained.

[0059] Among them, according to the information value I of each disaster control factor calculated, the ArcGIS raster calculator can be used to superimpose the I value to obtain the susceptibility index of each raster, and the Natural breaks method is used to divide the susceptibility of bedding landslides in the study area into four classification intervals: high, medium, low, and non, to obtain the susceptibility zoning map of bedding landslides in the study area. Please refer to Figure 6 Figure 6 A susceptibility zoning map of bedding landslides in the study area provided in the embodiments of the present application.

[0060] The specific implementation of step S6 may include: InSAR deformation monitoring is performed on the bedding slope in the high-susceptibility zone, and deformation time series data is obtained from the high-susceptibility zone in the susceptibility zone map.

[0061] InSAR deformation monitoring specifically includes: The interference pairs of the slopes along the bedding in the high-risk area are differentially processed, and the nonlinear deformation and noise phase are separated by spatiotemporal filtering. The deformation time series is reconstructed by combining the weighted least squares method and the temporal coherence coefficient, and the temporal coherence coefficient is used to screen the coherent points. For example, for areas prone to high and extremely high landslides along the layers, interference pairs within the appropriate space-time baseline threshold are selected. First, differential interference processing and phase unwrapping are implemented to obtain low-frequency deformation phase, terrain error phase, and terrain residual phase; then, the space-time filtering method is used to separate the nonlinear deformation, atmosphere, and noise phases in the residual phase, eliminate terrain errors, atmospheric delay errors, and noise, retain low-frequency deformations and nonlinear deformations, and obtain the error-corrected deformation data set; finally, the weighted least squares method is used to reconstruct the deformation time series, and the temporal coherence coefficient is used to screen high-quality coherent points to improve monitoring accuracy. Through this step, InSAR technology is used to quantitatively identify landslide hazards based on temporal and spatial characteristics such as deformation rate, cumulative deformation magnitude, and deformation area boundaries. Please refer to Figure 7 , Figure 7 A schematic diagram of the InSAR deformation rate provided in an embodiment of the present application.

[0062] Among them, based on the preset space-time baseline threshold (usually controlling the spatial baseline <200 meters and the time baseline <30 days to reduce the risk of loss of correlation), high-quality synthetic aperture radar (SAR) image interference pairs are screened to ensure data coherence; then DInSAR (Differential Interferometric Synthetic Aperture Radar) is implemented. DInSAR is a remote sensing method that uses the phase difference information of synthetic aperture radar (SAR) images to detect small surface deformations (such as earthquakes, volcanic activities, ground subsidence, etc.) through differential interferometric processing technology. The phase unwrapping is used to convert the entangled phase into an absolute phase difference, and the low-frequency deformation phase (reflecting slow landslide movement), the terrain error phase (caused by the error of the digital elevation model DEM) and the residual phase (including atmospheric delay, noise and nonlinear deformation) are separated. Then, the atmospheric disturbance (such as ionospheric delay, water vapor change) and random noise are removed from the residual phase by using spatiotemporal filtering technology (such as wavelet decomposition or principal component analysis), while retaining the low-frequency deformation (such as creep) and sudden nonlinear deformation (such as local collapse) related to the landslide, forming a high-confidence deformation data set after error correction. The weighted least squares method is further used to calculate the residual phase. The InSAR technology is used to fuse multi-phase SAR data with the WLS (Wall and Squares) method to reconstruct the surface deformation time series, and the stable coherent points are screened by the temporal coherence coefficient (γ>0.7), and low-quality pixels (such as vegetation coverage areas) are excluded. Finally, key parameters such as deformation rate (such as millimeter / year level), cumulative displacement (such as centimeter-level mutation) and deformation space boundary (such as the contour of continuous deformation area) are extracted. The abnormal deformation area is delineated in combination with the landslide susceptibility zoning, and the coordinated analysis of InSAR technology and slope structural characteristics is realized, so as to accurately identify the hidden dangers of bedding landslides in the creeping or acceleration stage.

[0063] The specific implementation of step S7 may include: For the landslide sites along the bedding, high-resolution images without cloud cover are selected to show the three-dimensional geomorphic features of the corresponding area. Through optical remote sensing (RS) image analysis, landslide hazard characteristics of the bedding slopes, such as drum hills at the front edge, river diversions, steep slopes behind the landslide, etc., are identified, and finally the sites with potential landslide hazards along the bedding in the area are determined. Figure 8 , Figure 8 A schematic diagram of the final determination of a bedding landslide provided in an embodiment of the present application.

[0064] In the above implementation process, by integrating multi-source remote sensing data with dynamic analysis of slope structure, accurate identification and graded early warning of bedding landslide hazards can be achieved. By quickly locking the bedding slopes based on terrain segmentation and rock formation screening, the spatial differences in susceptibility are quantified by combining the disaster control factor information model, and a closed-loop verification mechanism is formed by deformation monitoring and optical image feature verification. This effectively solves the problems of traditional methods relying on manual experience, ignoring the particularity of slope structure, and insufficient identification accuracy in complex mountainous environments. It improves the identification efficiency and accuracy of bedding landslide hazards, and provides data-driven technical support for engineering construction and disaster prevention and control in dangerous mountainous areas.

[0065] The embodiments described above 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 embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for identifying hidden dangers of landslides along the layer based on slope structure and multi-source remote sensing technology, characterized in that: include: Obtain elevation images and historical landslide data of the study area to divide the study area into units, and obtain each slope unit in the study area; In the slope unit, screening is performed according to the stratum inclination, the angle of slope direction and the number of free faces to obtain the bedding slope; Extracting the controlling factors of the landslide disaster of the bedding slope; Taking the disaster control factor as an input parameter of an information volume model, determining the information volume value of each of the disaster control factors based on the information volume model; Based on the grid calculator, all the information values ​​are superimposed to determine the susceptibility index of each grid, and the susceptibility index is divided into a plurality of classification intervals to obtain a susceptibility zoning map of the bedding landslide in the study area; Acquire deformation data from a partition with high susceptibility in the susceptibility partition map, and verify the partition result of the susceptibility partition based on the deformation data; The optical remote sensing images and the susceptibility zoning map are combined to determine the construction sites in the study area that have the risk of bedding landslide.

2. The method according to claim 1, characterized in that: The elevation image and historical landslide data of the study area are obtained to divide the study area into units, and each slope unit in the study area is obtained, including: Acquire an elevation image and historical landslide data of a study area, and extract a slope aspect and a mountain shadow map from the elevation image and the historical landslide data respectively; Determining parameters of a multi-scale segmentation algorithm based on the slope aspect and the hill shadow map; the parameters of the multi-scale segmentation algorithm include scale, shape feature weight and compactness weight parameters; Combining the modified trial and error method with the parameters of the multi-scale segmentation algorithm, each slope unit in the study area is obtained after all areas are merged through pixel aggregation and heterogeneity threshold merging.

3. The method according to claim 1, characterized in that In the slope unit, screening is performed according to the rock layer inclination, the angle of slope direction and the number of free faces to obtain the layer slope, including: Slope units whose angle between rock stratum dip and slope aspect is less than the set threshold are classified as forward slopes, and the threshold range of the angle is adjusted according to the conditions of single-sided empty space, double-sided empty space, and three-sided empty space.

4. The method according to claim 3, characterized in that The slope units whose angle between the stratum dip and the slope direction is less than the set threshold are classified as the forward slope, and the threshold range of the angle is adjusted according to the conditions of one side facing open, two sides facing open, and three sides facing open, including: The slope structure combination with an inclination angle less than 10 degrees is classified as a nearly horizontal structural slope; Generate interpolated rock formation dip by Kriging interpolation, and perform grid calculation on the interpolated rock formation dip and slope direction to obtain the angle between the rock formation dip and slope direction; For a single-sided open-air slope, slope work points with an angle less than 30 degrees are retained; for a double-sided open-air slope, slope work points with an angle less than 60 degrees are retained; for a three-sided open-air slope, slope work points with an angle less than 90 degrees are retained.

5. The method according to claim 1, characterized in that The step of extracting the controlling factors of the landslide disaster of the bedding slope comprises: The frequency ratios of topography, stratum lithology, geological structure, hydro-meteorological and engineering disturbance indicators in the bedding slope are calculated, and the five factors with the largest frequency ratios are selected as disaster control factors.

6. The method according to claim 1, characterized in that The method of using the disaster control factor as an input parameter of an information volume model and determining the information volume value of each disaster control factor based on the information volume model includes: Assume that each disaster control factor is X i ( i =1,2,3……), for landslide events H The amount of information provided I ( X i , H ) is calculated using the sample frequency form: in, N is the total number of landslide hazards in the study area; T is the total number of evaluation units in the study area; For the i The first of the three controlling factors j The number of landslides of the type; For the i Among the disaster-controlling factors j The number of evaluation units occupied by the class.

7. The method according to claim 1, characterized in that The grid-based calculator superimposes all the information values, determines the susceptibility index of each grid, divides the susceptibility index into a plurality of grading intervals, and obtains a susceptibility zoning map of the bedding landslide in the study area, including: The information value of each disaster control factor is superimposed based on the ArcGIS raster calculator to obtain the susceptibility index of each raster.

8. The method according to claim 7, characterized in that The grid-based calculator superimposes all the information values, determines the susceptibility index of each grid, divides the susceptibility index into a plurality of grading intervals, and obtains a susceptibility zoning map of the bedding landslide in the study area, and also includes: The natural breakpoint method was used to divide the susceptibility of bedding landslides in the study area into four classification intervals, and a susceptibility zoning map of bedding landslides in the study area was obtained.

9. The method according to claim 1, characterized in that: The step of acquiring deformation data from a partition with high susceptibility in the susceptibility partition map and verifying a partition result of the susceptibility partition based on the deformation data includes: InSAR deformation monitoring is performed on the bedding slope in the high-susceptibility zone, and deformation time series data is obtained from the high-susceptibility zone in the susceptibility zone map.

10. The method according to claim 9, characterized in that The InSAR deformation monitoring includes: The interference pairs of the slope along the bedding in the high-prone area are differentially processed, and the nonlinear deformation and noise phase are separated by spatiotemporal filtering. The deformation time series is reconstructed by combining the weighted least squares method and the temporal coherence coefficient, and the temporal coherence coefficient is used to screen the coherent points.

Images