Landslide identification method and device, electronic equipment and storage medium

Abstract

The application discloses a landslide identification method and device, electronic equipment and a storage medium, and belongs to the technical field of geological exploration. The landslide identification method comprises the following steps: acquiring landslide basic data and high-resolution remote sensing image data of a target area; extracting landslide attribute features according to the landslide basic data; determining a landslide high-occurrence area according to the landslide attribute features; determining corresponding remote sensing image grid data of the landslide high-occurrence area in the high-resolution remote sensing image data; inputting the remote sensing image grid data into a trained landslide identification model to obtain a landslide boundary. The method is based on medium and low resolution data and high resolution data, and combines a machine learning model to identify the boundary of a landslide in the target area, thereby greatly saving manpower and resources, improving the efficiency and accuracy of landslide identification, and enabling the area of the landslide to be calculated according to the boundary of the landslide.

Description

Technical Field

[0001] This application belongs to the field of geological exploration technology, specifically relating to a landslide identification method, device, electronic equipment, and storage medium. Background Technology

[0002] The steep terrain of high-altitude mountain climate zones provides favorable dynamic conditions for surface processes and disasters. Combined with frost heave in fissures within these areas, glacial melting easily leads to instability at the valley shoulders of glaciers, triggering secondary disasters such as landslides. These disasters seriously threaten the safety of infrastructure such as roads and regional sustainable development. Against the backdrop of global warming, glacier area is continuously decreasing, the snow line is rising, and glacier retreat and thinning are significantly increasing, leading to a growing risk of secondary disasters such as landslides and debris flows.

[0003] The frequent occurrence of landslides in high-altitude mountain climate zones and the resulting hazards have attracted attention, and extensive landslide surveys, monitoring, and prevention efforts have been carried out in such zones as the China-Pakistan Economic Corridor. However, landslides in high-altitude mountain climate zones are the result of the combined effects of multiple factors, and their disaster-prone environment is complex. How to efficiently monitor, warn against, and prevent landslide disasters triggered by the strong coupling of geology, geomorphology, and meteorology remains a major challenge.

[0004] The primary condition for studying landslides is landslide identification. Landslide hazards are characterized by their high degree of concealment and difficulty in observation. Traditional landslide detection methods require a large amount of manpower and resources and have limited applicability. With the rapid development of computer and remote sensing technologies, methods such as machine learning and deep learning have been applied to landslide identification. However, the accuracy of the results largely depends on the resolution of the image, and high-resolution images are difficult to acquire and data processing is time-consuming. Therefore, this method is more suitable for identifying landslides in small areas. Summary of the Invention

[0005] The purpose of this application is to provide a landslide identification method, device, electronic device, and storage medium to solve the problem that existing landslide identification methods require a large amount of manpower and material resources.

[0006] According to a first aspect of the embodiments of this application, a landslide identification method is provided, the method comprising:

[0007] Acquire basic landslide data and high-resolution remote sensing image data of the target area;

[0008] Extract landslide attribute features based on the aforementioned landslide basic data;

[0009] Determine high-incidence areas of landslides based on the aforementioned landslide attribute characteristics;

[0010] Determine the remote sensing image raster data corresponding to the landslide-prone areas in the high-resolution remote sensing image data;

[0011] The remote sensing image raster data is input into the trained landslide recognition model to obtain the landslide boundary.

[0012] In some optional embodiments of this application, the landslide basic data includes: the geographical location of the landslide, the boundary of the landslide, the DEM of the target area, the Landsat 8 image of the target area, the river and road vector map of the target area, the seismic data of the target area, the glacier distribution status of the target area, and the soil geological data of the target area.

[0013] In some optional embodiments of this application, the step of extracting landslide attribute features based on the landslide basic data includes:

[0014] Hydrological analysis was performed on the DEM of the target area to generate positive and negative catchment basins. The positive and negative catchment basins were then merged. After artificial modification of unreasonable units, slope units composed of valley lines and ridge lines were obtained.

[0015] Calculate the NVDI of the target region based on the Landsat 8 imagery of the target region;

[0016] Based on the DEM of the target area, calculate the slope, aspect and curvature of the target area;

[0017] Perform river and road buffer zone analysis based on the river and road vector maps of the target area;

[0018] Kernel density analysis was performed based on historical earthquake data of the target area.

[0019] Buffer analysis based on glacier distribution data in the target area;

[0020] Soil erodibility factors were calculated based on soil texture data of the target area;

[0021] To ensure consistent spatial resolution of landslide attribute characteristics;

[0022] By overlaying and analyzing the slope unit layer and each factor layer, the factor attribute characteristics of the slope unit are obtained.

[0023] By combining the landslide vector data of the target area with the factor attribute characteristics of the slope unit, the landslide attribute characteristics are obtained.

[0024] In some optional embodiments of this application, determining landslide-prone areas based on the landslide attribute characteristics includes:

[0025] The landslide attribute features are input into the trained landslide susceptibility identification model to obtain landslide-prone areas.

[0026] In some optional embodiments of this application, the trained landslide susceptibility identification model is trained using the following method:

[0027] Using the occurrence of landslides as the prediction target, a prediction model for the target variable is established, and the data is processed and fitted.

[0028] Standardize and transform the parameters of each factor.

[0029] The landslide sample data was randomly divided into training data and test data;

[0030] Perform multi-fold cross-validation on the training data;

[0031] By examining the scores of the training data through multi-fold cross-validation, the optimal parameters are selected to obtain a well-trained landslide susceptibility identification model.

[0032] In some optional embodiments of this application, the trained landslide identification model is obtained by the following method:

[0033] Using landslide vector data, a spatial buffer analysis with a distance of D is performed on each landslide to obtain the landslide outward polygon;

[0034] Intersection analysis is performed between each sample range in the sample range dataset and the landslide vector data to obtain the landslide polygon within each sample range;

[0035] Obtain remote sensing image raster data within the landslide sample area and adjust the image channels to serve as the image content for that sample;

[0036] Find the landslide mask polygons corresponding to the landslide sample range in the landslide mask dataset, and rasterize them as the mask data for that sample.

[0037] The obtained sample image content and sample mask data together constitute the landslide training samples, and all landslide samples are numbered to form a landslide sample library;

[0038] The landslide detection model is trained using the landslide sample library to obtain a trained landslide recognition model.

[0039] In some optional embodiments of this application, after obtaining the landslide boundary, the landslide identification method further includes:

[0040] Calculate the landslide area based on the landslide boundary.

[0041] According to a second aspect of the embodiments of this application, a landslide identification device is provided, the device may include:

[0042] The acquisition module is used to acquire basic landslide data and high-resolution remote sensing image data of the target area.

[0043] The extraction module is used to extract landslide attribute features based on the landslide basic data;

[0044] The high-incidence area determination module is used to determine high-incidence areas of landslides based on the landslide attribute characteristics.

[0045] The remote sensing image determination module is used to determine the remote sensing image raster data corresponding to the landslide-prone area in the high-resolution remote sensing image data;

[0046] The identification module is used to input the remote sensing image raster data into the trained landslide identification model to obtain the landslide boundary.

[0047] According to a third aspect of the embodiments of this application, an electronic device is provided, which may include:

[0048] processor;

[0049] Memory used to store processor-executable instructions;

[0050] The processor is configured to execute instructions to implement the landslide identification method as shown in any embodiment of the first aspect.

[0051] According to a fourth aspect of the embodiments of this application, a storage medium is provided, which, when the instructions in the storage medium are executed by a processor of an information processing device or a server, causes the information processing device or server to implement the landslide identification method as shown in any embodiment of the first aspect.

[0052] The above-mentioned technical solution of this application has the following beneficial technical effects:

[0053] The method described in this application identifies the boundaries of landslides within a target area by using low-to-medium resolution data (such as DEM, rainfall, soil properties, rivers, earthquakes, glaciers, NDVI, etc.) and high-resolution data (high-precision remote sensing images), combined with a trained machine learning model. This method greatly saves manpower and resources, achieves high accuracy in landslide identification, and can calculate the landslide area based on the landslide boundaries. Attached Figure Description

[0054] Figure 1 This is a flowchart of a landslide identification method in an exemplary embodiment of this application;

[0055] Figure 2 This is a flowchart of a landslide intelligent identification method in a specific embodiment of this application;

[0056] Figure 3 This is a landslide distribution map already existing in a specific embodiment of this application;

[0057] Figure 4 This is a spatial pattern diagram of each factor in a specific embodiment of this application;

[0058] Figure 5 This is a key landslide area map based on random forest in a specific embodiment of this application;

[0059] Figure 6 This is a map of the key landslide area in a specific embodiment of this application;

[0060] Figure 7 This is a landslide identification result obtained based on the Mask R-CNN model in a specific embodiment of this application;

[0061] Figure 8 This is a landslide boundary obtained based on the Mask R-CNN model in a specific embodiment of this application;

[0062] Figure 9 This is a schematic diagram of the electronic device structure in an exemplary embodiment of this application;

[0063] Figure 10 This is a schematic diagram of the hardware structure of an electronic device in an exemplary embodiment of this application. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to specific embodiments and accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of this application. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0065] The accompanying drawings illustrate layer structure diagrams according to embodiments of this application. These drawings are not to scale, and some details have been enlarged for clarity, and some details may have been omitted. The shapes of the various regions and layers shown in the drawings, as well as their relative sizes and positional relationships, are merely exemplary and may deviate from reality due to manufacturing tolerances or technical limitations. Furthermore, those skilled in the art can design regions / layers with different shapes, sizes, and relative positions as needed.

[0066] Obviously, the described embodiments are only a part of the embodiments of this application, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0067] In the description of this application, it should be noted that the terms "first", "second", and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0068] Furthermore, the technical features involved in the different embodiments of this application described below can be combined with each other as long as they do not conflict with each other.

[0069] The landslide identification method provided in this application will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0070] like Figure 1 As shown, in a first aspect of this application, a landslide identification method is provided, which may include:

[0071] Step S110: Acquire basic landslide data and high-resolution remote sensing image data of the target area;

[0072] Step S120: Extract landslide attribute features based on the landslide basic data;

[0073] Step S130: Determine high-incidence areas of landslides based on the landslide attribute characteristics;

[0074] Step S140: Determine the remote sensing image raster data corresponding to the landslide-prone area in the high-resolution remote sensing image data;

[0075] Step S150: Input the remote sensing image raster data into the trained landslide recognition model to obtain the landslide boundary.

[0076] This embodiment of the method identifies the boundaries of landslides within a target area by using low-to-medium resolution data (such as DEM, rainfall, soil properties, rivers, earthquakes, glaciers, NDVI, etc.) and high-resolution data (high-precision remote sensing images), combined with a trained machine learning model. This greatly saves manpower and resources, achieves high accuracy in landslide identification, and can calculate the landslide area based on the landslide boundaries.

[0077] To illustrate this more clearly, the steps described above will be explained in detail below:

[0078] The first step is S110: acquire basic landslide data and high-resolution remote sensing image data of the target area.

[0079] The basic data for landslides in this step includes the landslide's geographical location (latitude and longitude), landslide boundaries, target area DEM, Landsat 8 imagery, earthquake data, glacier data, soil properties, and other basic data.

[0080] Then comes step S120: extracting landslide attribute features based on the landslide basic data.

[0081] This step may include:

[0082] The target area DEM was hydrologically analyzed using ArcGIS software to generate positive and negative catchment basins. The positive and negative catchment basins were then merged. After artificially modifying unreasonable units, slope units composed of valley lines and ridge lines were obtained as slope units of the target area.

[0083] Based on the Landsat 8 imagery of the target region, the NVDI of the target region was calculated using ENVI software;

[0084] Based on the DEM of the target area, ArcGIS software is used to calculate the slope, aspect, and curvature of the target area.

[0085] Based on the vector map of rivers and roads in the target area, ArcGIS software was used to perform river and road buffer analysis.

[0086] Kernel density analysis was performed using ArcGIS software based on historical earthquake data of the target area.

[0087] Based on glacier distribution data in the target area, buffer analysis was performed using ArcGIS software.

[0088] Based on soil texture data of the target area, soil erodibility factors were calculated using ArcGIS software.

[0089] Obtain multi-year average rainfall data for the target area, and resample the data of factors such as elevation, slope, aspect, slope curvature, NDVI, multi-year average rainfall, earthquake density, distance from road, distance from river network, distance from glacier, and soil erodibility to ensure consistent spatial resolution.

[0090] Based on ArcGIS software, the spatial analysis module is used to obtain the factor attribute characteristics of each slope unit by overlaying and analyzing the slope unit layer and each factor layer.

[0091] By using methods such as visual interpretation and field verification, vector data of landslides were obtained. Combined with the factor attribute characteristics of each slope unit, the factor attribute characteristics of each landslide were obtained, and a landslide sample database was constructed.

[0092] The next step is step S130: Determine the high-incidence area of ​​landslides based on the landslide attribute characteristics.

[0093] This step involves using the occurrence of landslides as the prediction target, establishing a predictive model for the target variable, and using Python's sklearn module to process and fit the data. The parameters of each factor are standardized using methods such as mean squared error, mean removal, or variance normalization. Landslide sample data is randomly divided into training and testing data, with the training set comprising 80% of the sample and used to train the model, and the testing set comprising 20% ​​of the sample and used to test the model. Ten-fold cross-validation is performed on the training data: in Python, the sklearn module is used to divide the training data into 10 groups. Each time, 9 groups are used as the training set, and the remaining group is not used for model training. The model fitted to the training set is evaluated, and this process is repeated 10 times, equivalent to sampling without replacement. Cross-validation can evaluate the model's fit and avoid overfitting. The parameter of the number of trees in the forest (max-depth) is adjusted, and the score of the training data is viewed through 10-fold cross-validation. By adjusting the max-depth value to make the curve converge, the optimal max-depth parameter is finally selected as the optimal parameter, and a well-trained landslide prediction model is obtained. Using the well-trained landslide prediction model, inputting data on factors such as target area, elevation, slope, aspect, slope curvature, NDVI, multi-year average rainfall, earthquake density, distance from road, distance from river network, distance from glacier, and soil erodibility, the landslide susceptibility prediction result is obtained.

[0094] The next step is step S140: Determine the remote sensing image raster data corresponding to the landslide-prone area in the high-resolution remote sensing image data.

[0095] The landslide high-incidence area prediction results obtained in the previous step are vector data. In this step, a spatial buffer analysis with a distance of D is performed on each landslide in the landslide high-incidence area to obtain the landslide outer polygon. The minimum outer rectangle of the landslide outer polygon is obtained, which is the spatial range of the landslide. Then, the remote sensing image raster data corresponding to the spatial range of the landslide is obtained using the GDAL library.

[0096] Finally, in step S150: the remote sensing image raster data is input into the trained landslide recognition model to obtain the landslide boundary.

[0097] This step involves inputting the remote sensing image corresponding to the predicted landslide in the target area into the trained landslide recognition model to obtain the recognition result of the landslide boundary; then, the landslide is located, specifically by obtaining the latitude and longitude of the boundary pixels of the image where the landslide is located, calculating the coordinates of each boundary pixel in the image, and then mapping the boundary pixels to GPS coordinates to achieve GPS positioning of each landslide.

[0098] The well-trained landslide identification model can be obtained through the following steps:

[0099] Using landslide vector data, a spatial buffer analysis with a distance of D is performed on each landslide to obtain the landslide outer polygon. D is generally 5 meters. Then, the minimum outer rectangle of the landslide outer polygon is obtained as the spatial range of the landslide, i.e. the landslide sample range.

[0100] By performing an intersection analysis between each sample range in the sample range dataset and the landslide vector data, the landslide polygon inside each sample range is obtained, i.e., the landslide mask.

[0101] The remote sensing image raster data within the landslide sample area is obtained using the GDAL library, and the image channels are adjusted to serve as the image content of the sample. Then, the corresponding landslide mask polygon within the landslide sample area is found in the landslide mask dataset, and it is vectorized and rasterized to serve as the mask data of the sample. The obtained sample image content and sample mask data together constitute the landslide training sample, and all landslide samples are numbered to form the landslide sample library.

[0102] The Mask R-CNN model structure was implemented using the TensorFlow, Keras machine learning framework, and Python programming language. The landslide sample library was used as the training sample source for the Mask R-CNN model. During the training process, the model requested samples by obtaining the sample image content and sample landslide mask data through the sample number in the landslide sample library.

[0103] The Mask R-CNN model executable program, which uses a landslide sample library as the sample input source, is used to train the landslide detection model, resulting in a trained landslide recognition model.

[0104] In some optional embodiments of this application, the landslide basic data includes: the geographical location of the landslide, the boundary of the landslide, the DEM of the target area, the Landsat 8 image of the target area, the river and road vector map of the target area, the seismic data of the target area, the glacier distribution status of the target area, and the soil geological data of the target area.

[0105] In some optional embodiments of this application, the step of extracting landslide attribute features based on the landslide basic data includes:

[0106] Hydrological analysis was performed on the DEM of the target area to generate positive and negative catchment basins. The positive and negative catchment basins were then merged. After artificial modification of unreasonable units, slope units composed of valley lines and ridge lines were obtained.

[0107] Calculate the NVDI of the target region based on the Landsat 8 imagery of the target region;

[0108] Based on the DEM of the target area, calculate the slope, aspect and curvature of the target area;

[0109] Perform river and road buffer zone analysis based on the river and road vector maps of the target area;

[0110] Kernel density analysis was performed based on historical earthquake data of the target area.

[0111] Buffer analysis based on glacier distribution data in the target area;

[0112] Soil erodibility factors were calculated based on soil texture data of the target area;

[0113] To ensure consistent spatial resolution of landslide attribute characteristics;

[0114] By overlaying and analyzing the slope unit layer and each factor layer, the factor attribute characteristics of the slope unit are obtained.

[0115] By combining the landslide vector data of the target area with the factor attribute characteristics of the slope unit, the landslide attribute characteristics are obtained.

[0116] In some optional embodiments of this application, determining landslide-prone areas based on the landslide attribute characteristics includes:

[0117] The landslide attribute features are input into the trained landslide susceptibility identification model to obtain landslide-prone areas.

[0118] In some optional embodiments of this application, the trained landslide susceptibility identification model is trained using the following method:

[0119] Using the occurrence of landslides as the prediction target, a prediction model for the target variable is established, and the data is processed and fitted.

[0120] Standardize and transform the parameters of each factor.

[0121] The landslide sample data was randomly divided into training data and test data;

[0122] Perform multi-fold cross-validation on the training data;

[0123] By examining the scores of the training data through multi-fold cross-validation, the optimal parameters are selected to obtain a well-trained landslide susceptibility identification model.

[0124] In some optional embodiments of this application, the trained landslide identification model is obtained by the following method:

[0125] Using landslide vector data, a spatial buffer analysis with a distance of D is performed on each landslide to obtain the landslide outward polygon;

[0126] Intersection analysis is performed between each sample range in the sample range dataset and the landslide vector data to obtain the landslide polygon within each sample range;

[0127] Obtain remote sensing image raster data within the landslide sample area and adjust the image channels to serve as the image content for that sample;

[0128] Find the landslide mask polygons corresponding to the landslide sample range in the landslide mask dataset, and rasterize them as the mask data for that sample.

[0129] The obtained sample image content and sample mask data together constitute the landslide training samples, and all landslide samples are numbered to form a landslide sample library;

[0130] The landslide detection model is trained using the landslide sample library to obtain a trained landslide recognition model.

[0131] In some optional embodiments of this application, after obtaining the landslide boundary, the landslide identification method further includes:

[0132] Calculate the landslide area based on the landslide boundary.

[0133] In one specific embodiment of this application, a method for landslide identification over a large area is provided. Based on medium-resolution data (such as DEM, rainfall, soil properties, river, earthquake, glacier, NDVI, and other basic data) and high-resolution data (high-precision remote sensing imagery), combined with random forest and an improved Mask R-CNN method, the method identifies the boundaries of landslides within a target area, obtains the latitude and longitude of the landslide boundary pixels, locates each landslide, and calculates its area. This method improves the accuracy of intelligent landslide identification and makes landslide identification more efficient.

[0134] A flowchart of a landslide intelligent identification method based on random forest and an improved Mask R-CNN model is shown below. Figure 2 As shown, the landslide data obtained ( Figure 3 High-precision remote sensing data, DEM data, rainfall, soil properties, river, earthquake, glacier, and NDVI data are combined with random forest and an improved Mask R-CNN method to achieve intelligent identification of landslides in target areas. The process includes the following steps:

[0135] (1) Landslide attribute feature extraction

[0136] ① Based on the target area's DEM data, rainfall, soil properties, rivers, earthquakes, glaciers, Landsat 8 imagery, and other data, ArcGIS software was used to obtain data on factors such as slope units, multi-year average rainfall, elevation, slope, aspect, slope curvature, NDVI, multi-year average rainfall, earthquake density, distance from roads, distance from river networks, distance from glaciers, and soil erodibility. The data was then resampled to ensure consistent spatial resolution. Figure 4 );

[0137] ② Using the ArcGIS spatial analysis module, the factor attribute characteristics of each slope unit are obtained by overlaying and analyzing the slope unit layer and each factor layer.

[0138] ③ By visual interpretation and field verification, landslide vector data of the target area are obtained. Combined with the factor attribute characteristics of each slope unit, the factor attribute characteristics of each landslide are obtained, and a landslide sample database is constructed.

[0139] (2) Assessment of key landslide areas

[0140] ① Standardize and transform the parameters of each factor by means of variance, mean removal or variance normalization; randomly select landslide sample data and divide it into two parts: training data (80%) and test data (20%), and use random forest for model training;

[0141] ② In Python, the model is subjected to 10-fold cross-validation by calling the sklearn module. The parameter of the number of trees in the forest (max-depth) is adjusted. The scores of the training data are viewed through 10-fold cross-validation. Finally, the optimal max-depth parameter is selected as the optimal parameter to obtain the trained landslide prediction model.

[0142] ③ Using the trained landslide prediction model, input the data of each factor in the target area to obtain the prediction results of the regional landslide susceptibility. Areas with high landslide susceptibility are designated as key landslide areas. Figure 5 );

[0143] ④ Perform a spatial buffer analysis with a distance of D on each landslide within the key landslide area identified in the assessment to obtain the spatial extent of the landslide. Then, use the GDAL library to obtain the remote sensing image raster data corresponding to the spatial extent of the landslide.

[0144] (3) Landslide identification

[0145] ①Based on landslide vector sample data, a spatial buffer analysis with a distance of 5 meters is performed on each landslide, and then the minimum outer rectangle of the landslide outer polygon is obtained to obtain the landslide sample range;

[0146] ② Perform intersection analysis between each sample range in the sample range dataset and the landslide vector data to obtain the landslide polygon inside each sample range, i.e., the landslide mask;

[0147] ③ Use the GDAL library to obtain remote sensing image raster data within the landslide sample area as image content; find the corresponding landslide mask polygon in the landslide mask dataset and rasterize it as mask data; the sample image content and mask data together constitute landslide training samples, and number all landslide samples to form a landslide sample library;

[0148] ④ The Mask R-CNN model structure was implemented using the TensorFlow, Keras machine learning framework and Python programming language. The landslide sample library was used as the training sample for the Mask R-CNN model. During the training process, the model requested samples by obtaining the sample image content and sample landslide mask data through the sample number in the landslide sample library, thus obtaining the trained landslide recognition model.

[0149] ⑤ Input the remote sensing images of key landslide areas into the trained landslide recognition model to obtain the recognition results of landslide boundaries. Obtain the latitude and longitude of the boundary pixels of the landslides in the image, calculate the coordinates of the boundary pixels in the image, and then generate a shapefile of the landslide boundaries. This allows for the localization of each landslide and the calculation of its area. For demonstration purposes, a section of the China-Pakistan Economic Corridor is selected as a case study. Figure 6 For key landslide areas based on random forest, Figure 7 The landslide identification results are based on the Mask R-CNN model. Figure 8 This is the boundary of the landslide.

[0150] It should be noted that the landslide identification method provided in this application can be executed by a landslide identification device or a control module within that device for performing the landslide identification method. This application uses the landslide identification device performing the landslide identification method as an example to illustrate the landslide identification device provided in this application.

[0151] In a second aspect of this application, a landslide identification device is provided, which may include:

[0152] The acquisition module is used to acquire basic landslide data and high-resolution remote sensing image data of the target area.

[0153] The extraction module is used to extract landslide attribute features based on the landslide basic data;

[0154] The high-incidence area determination module is used to determine high-incidence areas of landslides based on the landslide attribute characteristics.

[0155] The remote sensing image determination module is used to determine the remote sensing image raster data corresponding to the landslide-prone area in the high-resolution remote sensing image data;

[0156] The identification module is used to input the remote sensing image raster data into the trained landslide identification model to obtain the landslide boundary.

[0157] The landslide identification device in this application embodiment can be a device, or a component, integrated circuit, or chip in a terminal. The device can be a mobile electronic device or a non-mobile electronic device. For example, mobile electronic devices can be mobile phones, tablets, laptops, PDAs, in-vehicle electronic devices, wearable devices, ultra-mobile personal computers (UMPCs), netbooks, or personal digital assistants (PDAs), etc., while non-mobile electronic devices can be servers, network-attached storage (NAS), personal computers (PCs), televisions (TVs), ATMs, or self-service machines, etc. This application embodiment does not impose specific limitations.

[0158] The landslide identification device in this application embodiment can be a device with an operating system. This operating system can be Android, iOS, or other possible operating systems; this application embodiment does not specifically limit the specific operating system used.

[0159] The landslide identification device provided in this application embodiment can achieve... Figure 1 The various processes implemented in the method implementation examples will not be described again here to avoid repetition.

[0160] Optionally, such as Figure 9 As shown, this application embodiment also provides an electronic device 900, including a processor 901, a memory 902, and a program or instructions stored in the memory 902 and executable on the processor 901. When the program or instructions are executed by the processor 901, they implement the various processes of the above-described landslide identification method embodiment and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0161] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.

[0162] Figure 10 A schematic diagram of the hardware structure of an electronic device to implement an embodiment of this application.

[0163] The electronic device 1000 includes, but is not limited to, components such as: radio frequency unit 1001, network module 1002, audio output unit 1003, input unit 1004, sensor 1005, display unit 1006, user input unit 1007, interface unit 1008, memory 1009, and processor 1010.

[0164] Those skilled in the art will understand that the electronic device 1000 may also include a power supply (such as a battery) for supplying power to various components. The power supply may be logically connected to the processor 1010 through a power management system, thereby enabling functions such as managing charging, discharging, and power consumption through the power management system. Figure 10 The electronic device structure shown does not constitute a limitation on the electronic device. The electronic device may include more or fewer components than shown, or combine certain components, or have different component arrangements, which will not be elaborated here.

[0165] It should be understood that, in this embodiment, the input unit 1004 may include a graphics processing unit (GPU) 10041 and a microphone 10042. The GPU 10041 processes image data of still images or videos obtained by an image capture device (such as a camera) in video capture mode or image capture mode. The display unit 1006 may include a display panel 10061, which may be configured in the form of a liquid crystal display, an organic light-emitting diode, etc. The user input unit 1007 includes a touch panel 10071 and other input devices 10072. The touch panel 10071 is also called a touch screen. The touch panel 10071 may include a touch detection device and a touch controller. Other input devices 10072 may include, but are not limited to, physical keyboards, function keys (such as volume control buttons, power buttons, etc.), trackballs, mice, joysticks, etc., which will not be described in detail here. The memory 1009 can be used to store software programs and various data, including but not limited to applications and operating systems. Processor 1010 can integrate an application processor and a modem processor. The application processor mainly handles the operating system, user interface, and applications, while the modem processor mainly handles wireless communication. It is understood that the modem processor may also not be integrated into processor 1010.

[0166] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described landslide identification method embodiments and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0167] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0168] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above-described landslide identification method embodiments and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0169] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0170] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0171] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0172] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A landslide identification method, characterized in that, include: Acquire basic landslide data and high-resolution remote sensing image data of the target area; Extract landslide attribute features based on the aforementioned landslide basic data; Determine high-incidence areas of landslides based on the aforementioned landslide attribute characteristics; Determine the remote sensing image raster data corresponding to the landslide-prone areas in the high-resolution remote sensing image data; The remote sensing image raster data is input into the trained landslide recognition model to obtain the landslide boundary; The trained landslide identification model was obtained through the following method: Using landslide vector data, a spatial buffer analysis with a distance of D is performed on each landslide to obtain the landslide outward polygon; Intersection analysis is performed between each sample range in the sample range dataset and the landslide vector data to obtain the landslide polygon within each sample range; Obtain remote sensing image raster data within the landslide sample area and adjust the image channels to serve as the image content for that sample; Find the landslide mask polygons corresponding to the landslide sample range in the landslide mask dataset, and rasterize them as the mask data for that sample. The obtained sample image content and sample mask data together constitute the landslide training samples, and all landslide samples are numbered to form a landslide sample library; The landslide detection model is trained using the landslide sample library to obtain a trained landslide recognition model.

2. The landslide identification method according to claim 1, characterized in that, The landslide baseline data includes: the geographical location of historical landslides, the boundaries of historical landslides, the DEM of the target area, Landsat 8 imagery of the target area, vector maps of rivers and roads in the target area, seismic data of the target area, glacier distribution status of the target area, and soil geological data of the target area.

3. The landslide identification method according to claim 2, characterized in that, The step of extracting landslide attribute features based on the landslide basic data includes: Hydrological analysis was performed on the DEM of the target area to generate positive and negative catchment basins. The positive and negative catchment basins were then merged. After artificial modification of unreasonable units, slope units composed of valley lines and ridge lines were obtained. Calculate the NVDI of the target region based on the Landsat 8 imagery of the target region; Based on the DEM of the target area, calculate the slope, aspect and curvature of the target area; Perform river and road buffer zone analysis based on the river and road vector maps of the target area; Kernel density analysis was performed based on historical earthquake data of the target area. Buffer analysis based on glacier distribution data in the target area; Soil erodibility factors were calculated based on soil texture data of the target area; To ensure consistent spatial resolution of landslide attribute characteristics; By overlaying and analyzing the slope unit layer and each factor layer, the factor attribute characteristics of the slope unit are obtained. By combining the landslide vector data of the target area with the factor attribute characteristics of the slope unit, the landslide attribute characteristics are obtained.

4. The landslide identification method according to claim 1, characterized in that, The process of determining high-risk landslide areas based on the landslide attribute characteristics includes: The landslide attribute features are input into the trained landslide susceptibility identification model to obtain landslide-prone areas.

5. The landslide identification method according to claim 4, characterized in that, The trained landslide susceptibility identification model was obtained through the following method: Using the occurrence of landslides as the prediction target, a prediction model for the target variable is established, and the data is processed and fitted. Standardize and transform the parameters of each factor. The landslide sample data was randomly divided into training data and test data; Perform multi-fold cross-validation on the training data; By examining the scores of the training data through multi-fold cross-validation, the optimal parameters are selected to obtain a well-trained landslide susceptibility identification model.

6. The landslide identification method according to claim 1, characterized in that, After obtaining the landslide boundary, the landslide identification method further includes: Calculate the landslide area based on the landslide boundary.

7. A landslide identification device, characterized in that, include: The acquisition module is used to acquire basic landslide data and high-resolution remote sensing image data of the target area. The extraction module is used to extract landslide attribute features based on the landslide basic data; The high-incidence area determination module is used to determine high-incidence areas of landslides based on the landslide attribute characteristics. The remote sensing image determination module is used to determine the remote sensing image raster data corresponding to the landslide-prone area in the high-resolution remote sensing image data; The identification module is used to input the remote sensing image raster data into the trained landslide identification model to obtain the landslide boundary; The training module is used to train the trained landslide recognition model using the following method: Using landslide vector data, a spatial buffer analysis with a distance of D is performed on each landslide to obtain the landslide outward polygon; Intersection analysis is performed between each sample range in the sample range dataset and the landslide vector data to obtain the landslide polygon within each sample range; Obtain remote sensing image raster data within the landslide sample area and adjust the image channels to serve as the image content for that sample; Find the landslide mask polygons corresponding to the landslide sample range in the landslide mask dataset, and rasterize them as the mask data for that sample. The obtained sample image content and sample mask data together constitute the landslide training samples, and all landslide samples are numbered to form a landslide sample library; The landslide detection model is trained using the landslide sample library to obtain a trained landslide recognition model.

8. An electronic device, characterized in that, include: A processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the landslide identification method as described in any one of claims 1-6.

9. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the landslide identification method as described in any one of claims 1-6.

Images