Landslide surface deformation monitoring method, system, equipment and medium

Abstract

The invention provides a landslide surface deformation monitoring method, system and device and a medium, and belongs to the technical field of landslide monitoring, and the method comprises the steps: firstly determining the boundary of a target landslide, laying a first identification point outside the boundary of the landslide, and laying a second identification point with a number in the boundary; a plurality of unmanned aerial vehicle itinerant paths are set based on the first identification point, a plurality of images of the target landslide are collected, and the three-dimensional coordinates of each point location are calculated; comparing the three-dimensional coordinate change of the second identification point in two adjacent acquisition periods, and calculating the coordinate change quantity of the second identification point; and finally, determining a surface deformation result of the target landslide according to the coordinate variable quantity, arranging a first identification point outside the landslide boundary to realize high-precision positioning of each point location of the target landslide, setting a tour path of the unmanned aerial vehicle according to the first identification point, and monitoring a second identification point in the landslide boundary. Therefore, the omnibearing monitoring of the deformation of the target landslide is realized.

Description

Technical Field

[0001] The present application relates to the technical field of landslide monitoring, and in particular to a method, system, equipment and medium for monitoring landslide surface deformation. Background Art

[0002] With the frequent occurrence of geological disasters, the demand for landslide monitoring is increasing. At present, landslide surface deformation monitoring mainly relies on the deployment of monitoring instruments, SAR imaging InSAR technology, optical image pixel migration technology and lidar elevation difference method. The deployment of monitoring instruments has high precision and time continuity, but is only applicable to local areas; InSAR and pixel migration technology are suitable for large-scale monitoring, but are limited by two-dimensional deformation data, large time intervals and high costs; LiDAR elevation difference method is flexible in operation, but can only monitor single-dimensional deformation in the elevation direction. These technologies have their own advantages and disadvantages, and it is difficult to fully meet the needs of landslide monitoring.

[0003] However, landslides may cause significant property losses, traffic disruptions and even casualties. More comprehensive and accurate monitoring methods are urgently needed to improve disaster warning capabilities and reduce hazards. Summary of the invention

[0004] In view of the above problems, embodiments of the present application provide a landslide surface deformation monitoring method, system, device and medium to overcome the above problems or at least partially solve the above problems.

[0005] In a first aspect, the present application provides a method for monitoring landslide surface deformation, the method comprising: Determine the landslide boundary of the target landslide to be monitored; Arranging a plurality of first identification points outside the landslide boundary of the target landslide, and arranging a plurality of second identification points inside the landslide boundary of the target landslide, wherein each of the second identification points is provided with a number identification; Based on the plurality of first identification points, a plurality of patrol paths are set, and a plurality of images of the target landslide are respectively collected by the drone according to the plurality of patrol paths, and based on the collected plurality of images, the three-dimensional coordinates of each point in the target landslide are obtained; According to the three-dimensional coordinates of each point in the target landslide, the three-dimensional coordinates of a plurality of the second identification points in each acquisition cycle in two adjacent acquisition cycles are determined respectively, wherein one acquisition cycle is when the drone completes one image acquisition along all patrol paths; According to the three-dimensional coordinates of a plurality of the second identification points in each of the two adjacent acquisition cycles, respectively determining a coordinate change amount of the three-dimensional coordinate of each of the second identification points in the two adjacent acquisition cycles; The surface deformation result of the target landslide is determined according to the coordinate change of each of the second identification points.

[0006] Optionally, the step of arranging a plurality of first identification points outside a landslide boundary of the target landslide includes: Acquiring a panoramic image of the target landslide; Based on the panoramic image, construct a three-dimensional spatial model of the target landslide; Taking the landslide boundary of the three-dimensional space model as a reference, determining the landslide centroid of the target landslide in the three-dimensional space model; Using the centroid of the landslide as a reference point, calibrating a plurality of measurement reference points outside the landslide boundary in the three-dimensional space model at preset interval angles; With reference to the plurality of measurement reference points in the three-dimensional space model, a plurality of the first identification points are arranged outside the landslide boundary of the target landslide.

[0007] Optionally, the arranging a plurality of second identification points within the landslide boundary of the target landslide includes: Acquiring a panoramic image of the target landslide; Based on the panoramic image, construct a three-dimensional spatial model of the target landslide; Based on the landslide evolution mechanism theory, the deformation velocity of each point of the target landslide is deduced in the three-dimensional space model; Based on the deformation speed of each of the points, the target landslide is divided into a plurality of sub-areas; According to the deformation speed of the points corresponding to each of the sub-areas, a plurality of deformation monitoring points are respectively calibrated for each of the sub-areas in the three-dimensional space model; With reference to the plurality of deformation monitoring points in the three-dimensional space model, a plurality of the second identification points are arranged within the landslide boundary of the target landslide.

[0008] Optionally, determining the coordinate change of the three-dimensional coordinate of each second identification point in the two adjacent acquisition cycles according to the three-dimensional coordinates of a plurality of second identification points in each acquisition cycle in the two adjacent acquisition cycles includes: Based on the three-dimensional coordinates of each of the second identification points in the two adjacent acquisition cycles, respectively determine the three-dimensional coordinates of the second identification point corresponding to each number identification in the two adjacent acquisition cycles; According to the three-dimensional coordinates of the second identification point corresponding to each numbering identification in the two adjacent acquisition cycles, the coordinate change amount of the second identification point corresponding to each numbering identification in the two adjacent acquisition cycles is determined.

[0009] Optionally, determining the surface deformation result of the target landslide according to the coordinate change of each second identification point includes: Determine the surface deformation result of the second identification point corresponding to each number identification according to the coordinate change amount of the second identification point corresponding to each number identification within the two adjacent acquisition cycles; According to the surface deformation result of the second identification point corresponding to each serial identification, the surface deformation result of the target landslide is obtained by multi-point interpolation calculation.

[0010] Optionally, before determining the three-dimensional coordinates of the second identification point corresponding to each serial identification in the two adjacent acquisition cycles, the method further includes: Performing image segmentation on the collected multiple images of the target landslide to obtain multiple first images including the first identification points and multiple second images including the second identification points; Based on the first image and the second image, a deep learning algorithm is used to train an image recognition model to identify the first identification point and the second identification point through the image recognition model.

[0011] Optionally, obtaining the three-dimensional coordinates of each point in the target landslide based on the collected multiple images includes: Performing pixel registration and pixel stitching on the multiple images collected to obtain a pixel-level composite image; Establishing a three-dimensional coordinate system based on any one of the first identification points in the pixel-level synthetic image; The three-dimensional coordinates of each point in the target landslide are determined according to the three-dimensional coordinate system.

[0012] Optionally, after determining the surface deformation result of the target landslide, the method further includes: Obtaining surface deformation results corresponding to the target landslide in multiple groups of acquisition cycles, wherein a group of acquisition cycles includes two adjacent acquisition cycles; The plurality of surface deformation results are fused to obtain a continuous deformation field of the target landslide within the plurality of acquisition cycles.

[0013] Optionally, after determining the surface deformation result of the target landslide, the method further includes: Obtaining correction values ​​of deformation results of the target landslide collected by a plurality of total station measuring points pre-arranged within the landslide boundary of the target landslide; The surface deformation result of the target landslide is corrected according to the multiple correction values.

[0014] In a second aspect of the present application, a landslide surface deformation monitoring system is provided, the system comprising: A determination module, used for determining the landslide boundary of the target landslide to be monitored; A layout module, used for arranging a plurality of first identification points outside the landslide boundary of the target landslide, and arranging a plurality of second identification points inside the landslide boundary of the target landslide, wherein each of the second identification points is provided with a number identification; A setting module, used to set a plurality of patrol paths based on the plurality of first identification points, collect a plurality of images of the target landslide by the drone according to the plurality of patrol paths, and obtain the three-dimensional coordinates of each point in the target landslide based on the collected plurality of images; A first determination module is used to determine the three-dimensional coordinates of a plurality of the second identification points in each acquisition cycle in two adjacent acquisition cycles according to the three-dimensional coordinates of each point in the target landslide, wherein one acquisition cycle is when the drone completes one image acquisition along all patrol paths; A second determination module, configured to determine, based on the three-dimensional coordinates of a plurality of the second identification points in each of the two adjacent acquisition cycles, a coordinate change amount of the three-dimensional coordinates of each of the second identification points in the two adjacent acquisition cycles; The third determination module is used to determine the surface deformation result of the target landslide according to the coordinate change of each of the second identification points.

[0015] Optionally, the arranging of a plurality of first identification points outside the landslide boundary of the target landslide, the arranging module comprises: A first acquisition submodule is used to acquire a panoramic image of the target landslide; A first construction submodule is used to construct a three-dimensional spatial model of the target landslide based on the panoramic image; A first determination submodule is used to determine the landslide centroid of the target landslide in the three-dimensional space model by taking the landslide boundary of the three-dimensional space model as a reference; A first calibration submodule is used to calibrate a plurality of measurement reference points outside the landslide boundary in the three-dimensional space model at preset interval angles, using the landslide centroid as a reference point; The first layout submodule is used to layout a plurality of the first identification points outside the landslide boundary of the target landslide with reference to the plurality of measurement reference points in the three-dimensional space model.

[0016] Optionally, the arranging of a plurality of second identification points within the landslide boundary of the target landslide, the arranging module comprises: A second acquisition submodule is used to acquire a panoramic image of the target landslide; A second construction submodule is used to construct a three-dimensional spatial model of the target landslide based on the panoramic image; A deduction submodule, for deducing the deformation velocity of each point of the target landslide in the three-dimensional space model based on the evolution mechanism theory of the landslide; A region division submodule, used for dividing the target landslide into a plurality of sub-regions based on the deformation velocity of each of the points; A second calibration submodule is used to calibrate a plurality of deformation monitoring points for each of the sub-areas in the three-dimensional space model according to the deformation speed of the points corresponding to each of the sub-areas; The second layout submodule is used to layout a plurality of the second identification points within the landslide boundary of the target landslide with reference to the plurality of the deformation monitoring points in the three-dimensional space model.

[0017] Optionally, the coordinate change amount of the three-dimensional coordinate of each second identification point in the two adjacent acquisition cycles is determined respectively according to the three-dimensional coordinates of a plurality of second identification points in each acquisition cycle in the two adjacent acquisition cycles, and the second determination module includes: A second determination submodule, configured to determine the three-dimensional coordinates of the second identification point corresponding to each numbering identification in the two adjacent acquisition periods based on the three-dimensional coordinates of each of the second identification points in the two adjacent acquisition periods; The third determination submodule is used to determine the coordinate change of the second identification point corresponding to each numbering identification within the two adjacent acquisition cycles according to the three-dimensional coordinates of the second identification point corresponding to each numbering identification within the two adjacent acquisition cycles.

[0018] Optionally, the third determining module includes: A fourth determination submodule, configured to determine the surface deformation result of the second identification point corresponding to each number identification according to the coordinate change amount of the second identification point corresponding to each number identification within the two adjacent acquisition cycles; The calculation submodule is used to obtain the surface deformation result of the target landslide through multi-point interpolation calculation according to the surface deformation result of the second identification point corresponding to each serial identification.

[0019] Optionally, the system further comprises: An image segmentation submodule, used for performing image segmentation on the multiple images of the target landslide collected to obtain multiple first images containing the first identification points and multiple second images containing the second identification points; An image recognition submodule is used to train an image recognition model using a deep learning algorithm based on the first image and the second image, so as to identify the first identification point and the second identification point through the image recognition model.

[0020] Optionally, the three-dimensional coordinates of each point in the target landslide are obtained based on the collected multiple images, and the setting module includes: A registration and stitching submodule, used for performing pixel registration and pixel stitching on the multiple images collected to obtain a pixel-level composite image; An establishing submodule, used for establishing a three-dimensional coordinate system based on any one of the first identification points in the pixel-level synthetic image as a reference; The fifth determination submodule is used to determine the three-dimensional coordinates of each point in the target landslide according to the three-dimensional coordinate system.

[0021] Optionally, the system further comprises: A third acquisition submodule is used to acquire surface deformation results corresponding to the target landslide in multiple groups of acquisition cycles, wherein a group of acquisition cycles includes two adjacent acquisition cycles; The fusion submodule is used to fuse the multiple surface deformation results to obtain the continuous deformation field of the target landslide within the multiple groups of acquisition cycles.

[0022] Optionally, the system further comprises: A fourth acquisition submodule is used to acquire a correction value of a deformation result of the target landslide collected by each of a plurality of total station measuring points pre-arranged within the landslide boundary of the target landslide; The correction submodule is used to correct the surface deformation result of the target landslide according to the multiple correction values.

[0023] In a third aspect of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the landslide surface deformation monitoring method as described in the first aspect of the present application.

[0024] In a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the landslide surface deformation monitoring method as described in the first aspect of the present application is implemented.

[0025] Beneficial effects of this application: The present application provides a method for monitoring deformation of a landslide surface, the method comprising: determining a landslide boundary of a target landslide to be monitored; arranging a plurality of first identification points outside the landslide boundary of the target landslide, and arranging a plurality of second identification points within the landslide boundary of the target landslide, wherein each of the second identification points is provided with a number identification; based on the plurality of first identification points, setting a plurality of patrol paths, collecting a plurality of images of the target landslide by means of a drone according to the plurality of patrol paths, and obtaining a three-dimensional coordinate of each point in the target landslide based on the collected plurality of images; determining the three-dimensional coordinates of a plurality of second identification points in each of two adjacent acquisition cycles according to the three-dimensional coordinates of each point in the target landslide; determining the coordinate change amount of the three-dimensional coordinate of each of the second identification points in the two adjacent acquisition cycles according to the three-dimensional coordinates of the plurality of second identification points in each of the two adjacent acquisition cycles; determining the surface deformation result of the target landslide according to the coordinate change amount of each of the second identification points.

[0026] The present application first determines the boundary of the target landslide, and arranges a first identification point outside the boundary of the landslide and a second identification point with a number inside the boundary; then, multiple drone patrol paths are set based on the first identification point, multiple images of the target landslide are collected and the three-dimensional coordinates of each point are calculated; then, the three-dimensional coordinate changes of the second identification point in two adjacent collection cycles are compared, and the coordinate change amount is calculated; finally, the surface deformation result of the target landslide is determined according to the coordinate change amount, and the first identification point is arranged outside the boundary of the landslide to achieve high-precision positioning of each point of the target landslide, and the drone patrol path is set according to the first identification point to monitor the second identification point within the boundary of the landslide, so as to achieve all-round monitoring of the deformation of the target landslide. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments of the present application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0028] Figure 1 It is a schematic diagram of the steps of a method for monitoring landslide surface deformation provided in an embodiment of the present application; Figure 2 is a schematic diagram of a landslide surface deformation monitoring system provided in an embodiment of the present application; Figure 3 It is a schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0029] The exemplary embodiments of the present application will be described in more detail below in conjunction with the accompanying drawings in the embodiments of the present application. Although the exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present application and to enable the scope of the present application to be fully communicated to those skilled in the art.

[0030] Based on the above problems, in a first aspect, the present application provides a method for monitoring landslide surface deformation, such as Figure 1 As shown, the method includes: Step S101, determining the landslide boundary of the target landslide to be monitored.

[0031] In this step, it is first necessary to determine the landslide boundary of the target landslide to be monitored. In this application, the landslide boundary of the target landslide can be determined by ground survey, remote sensing interpretation and other methods. In practical applications, the delineation of landslide boundaries not only depends on ground survey and remote sensing data, but also needs to be combined with actual terrain and geological conditions. According to the shape and scale of the landslide, remote sensing technology is used to accurately locate the boundary. Especially when delineating the boundary, it is necessary to avoid the deviation of the boundary due to complex terrain or data error to ensure the reliability of the monitoring data.

[0032] Step S102, arranging a plurality of first identification points outside the landslide boundary of the target landslide, and arranging a plurality of second identification points inside the landslide boundary of the target landslide, wherein each of the second identification points is provided with a number identification.

[0033] In this step, a plurality of first identification points are arranged outside the landslide boundary of the target landslide, a plurality of second identification points are arranged inside the boundary, and each second identification point is provided with a number identification.

[0034] In this application, the first identification points are mainly arranged in stable areas outside the landslide boundary. These areas are usually not affected by landslide deformation and are generally used to provide stable reference coordinates. The selection of these identification points should take into account geological stability, unobstructed vision and accessibility of drone photography to ensure the long-term stability and accuracy of the identification points. Usually, a location with relatively flat terrain and far away from the landslide activity area is selected to avoid the impact of deformation on these identification points.

[0035] The second identification point is set in the deformation area within the landslide boundary to monitor the deformation of the landslide surface. Due to the complexity of landslide deformation, the second identification point should be set in the area where large deformation may occur in order to capture the deformation trend of the landslide. Each identification point is marked with a number to help accurately identify each monitoring point and its data to avoid confusion caused by too many points.

[0036] In one embodiment, the identification points can be laid out in a variety of ways, including precise placement by drones, manual calibration, etc. The drone uses a high-precision positioning system to accurately place the identification points at predetermined locations. These points are fixed to the ground in different ways, such as using glue, concrete and other materials to ensure long-term stability and not be affected by the external environment. In addition, the layout of the identification points should ensure that a uniform network structure is formed in the landslide area. The layout of the first identification point needs to take into account the stable area outside the landslide, which is usually sparse. The layout of the second identification point should be based on the actual situation of the landslide, especially in areas with more significant deformation, and the layout density should be appropriately increased to improve monitoring accuracy.

[0037] Step S103, based on the multiple first identification points, multiple patrol paths are set, and multiple images of the target landslide are respectively collected by the drone according to the multiple patrol paths, and the three-dimensional coordinates of each point in the target landslide are obtained based on the collected multiple images.

[0038] In this step, multiple patrol paths are set based on multiple first identification points. In practical applications, each patrol path covers at least one first identification point. Since the layout of the first identification points takes into account the geographical environment and regional scale of the target landslide area, the setting of the patrol path should ensure that the entire target landslide and its boundaries are covered, especially the areas with more significant deformation. The path design needs to take into account the terrain undulations, obstacles, and the optimal angle and route of the drone flight to ensure the comprehensiveness and accuracy of subsequent image acquisition.

[0039] Optionally, the patrol path should not only cover the target landslide, but also comprehensively consider the flight altitude, angle, speed, shooting interval and other factors of the drone. The optimization of the patrol path should ensure that each first identification point can be fully photographed at multiple angles, thereby improving the accuracy of the subsequent reconstruction of the three-dimensional coordinates based on the first identification point.

[0040] Drones equipped with high-precision cameras fly regularly along patrol routes and collect image data in real time. By shooting from multiple angles and perspectives, images of landslides can be obtained from different directions and heights to ensure the comprehensiveness of the data.

[0041] The collected images are reconstructed in three dimensions through computer vision technology (such as stereo vision, Structure from Motion (SfM), multi-view geometry, etc.). The coordinates (X, Y, Z) of each first identification point in three-dimensional space are calculated through the perspective difference between the images, and a three-dimensional coordinate system is established with any first identification point as a reference, thereby obtaining the three-dimensional coordinates of each point of the target landslide. In this application, the point of the target landslide can be a pixel point of the collected image of the target landslide.

[0042] Step S104, according to the three-dimensional coordinates of each point in the target landslide, respectively determine the three-dimensional coordinates of multiple second identification points in each acquisition cycle within two adjacent acquisition cycles, wherein one acquisition cycle is when the drone completes one image acquisition along all patrol paths.

[0043] In this step, "collection cycle" refers to a complete image collection process performed by the drone according to all patrol paths. In each collection cycle, the drone will complete an image collection according to all patrol paths, and each image collection will collect multiple second identification points. Then, based on the three-dimensional coordinates of each point of the target landslide determined in the previous step, the three-dimensional coordinates of multiple second identification points in each collection cycle within two adjacent collections are determined.

[0044] Step S105, based on the three-dimensional coordinates of a plurality of the second identification points in each of the two adjacent acquisition cycles, respectively determine the coordinate change of the three-dimensional coordinates of each of the second identification points in the two adjacent acquisition cycles.

[0045] In this step, by comparing the three-dimensional coordinates of multiple second identification points in two adjacent acquisition cycles, the coordinate change of each second identification point between the two acquisition cycles is calculated. Specifically, the coordinate change refers to the difference in the three-dimensional coordinates (X, Y, Z) of the second identification point in two adjacent acquisition cycles. By comparing the three-dimensional coordinates of each second identification point in two acquisition cycles, the change in the position of the point in space can be obtained. For example, if between periods t1 and t2, the three-dimensional coordinates of the second identification point A are and , then the coordinate change of the point , , .

[0046] Step S106, determining the surface deformation result of the target landslide according to the coordinate change of each of the second identification points.

[0047] Surface deformation refers to the spatial displacement of the surface of the landslide body due to landslide activity or external factors within a certain period of time. By calculating the coordinate changes of each second identification point, the deformation information of the landslide surface in different areas can be obtained, and then the overall deformation of the landslide can be described. Since step S105 has calculated the coordinate changes of each second identification point between two adjacent acquisition cycles, the surface deformation of the target landslide can be further determined by these changes. In practical applications, the surface deformation result of the target landslide can be obtained by summarizing, weighted averaging or interpolating the coordinate changes of all second identification points. The surface deformation result of the target landslide is usually expressed as a deformation field or displacement field on the surface of the target landslide.

[0048] The present application first determines the boundary of the target landslide, and arranges a first identification point outside the boundary of the landslide and a second identification point with a number inside the boundary; then, multiple drone patrol paths are set based on the first identification point, multiple images of the target landslide are collected and the three-dimensional coordinates of each point are calculated; then, the three-dimensional coordinate changes of the second identification point in two adjacent collection cycles are compared, and the coordinate change amount is calculated; finally, the surface deformation result of the target landslide is determined according to the coordinate change amount, and the first identification point is arranged outside the boundary of the landslide to achieve high-precision positioning of each point of the target landslide, and the drone patrol path is set according to the first identification point to monitor the second identification point within the boundary of the landslide, so as to achieve all-round monitoring of the deformation of the target landslide.

[0049] In a preferred embodiment, a plurality of first identification points are arranged outside the landslide boundary of the target landslide, including: Acquiring a panoramic image of the target landslide; Based on the panoramic image, construct a three-dimensional spatial model of the target landslide; Taking the landslide boundary of the three-dimensional space model as a reference, determining the landslide centroid of the target landslide in the three-dimensional space model; Using the centroid of the landslide as a reference point, calibrating a plurality of measurement reference points outside the landslide boundary in the three-dimensional space model at preset interval angles; With reference to the plurality of measurement reference points in the three-dimensional space model, a plurality of the first identification points are arranged outside the landslide boundary of the target landslide.

[0050] In this embodiment, the panoramic image refers to a wide-angle view stitched together from high-resolution images taken from multiple angles and fields of view, which can fully display the topographic features of the landslide area. The purpose of obtaining the panoramic image is to provide sufficient image data for the subsequent construction of a three-dimensional space model. In practical applications, the panoramic image can be taken by a drone equipped with a high-definition camera, using multiple viewing angles and appropriate overlap for image acquisition. The flight path and shooting angle of the drone need to be designed according to the specific morphology and terrain conditions of the landslide to ensure that the image coverage is comprehensive and there are no blind spots.

[0051] By applying stereoscopic vision processing technology to panoramic images, three-dimensional spatial coordinates can be extracted from image data from multiple perspectives to construct a three-dimensional spatial model of the target landslide area. The three-dimensional model not only shows the surface morphology of the landslide, but also accurately reflects the terrain changes, providing basic data for subsequent landslide monitoring.

[0052] The landslide boundary of the three-dimensional spatial model is used as a reference to determine the landslide centroid of the target landslide in the three-dimensional spatial model. For example, the landslide centroid of the target landslide is determined by analyzing the landslide boundary and terrain characteristics using a geometric method or a weighted average method. The landslide centroid refers to the geometric center of the landslide body, which is usually the average center of gravity in the landslide area. In the three-dimensional spatial model, the determination of the landslide centroid provides a unified reference point for the layout of subsequent identification points, which helps to ensure the accurate positioning of the identification points.

[0053] Once the centroid of the landslide is located, the measurement benchmark points outside the landslide boundary are calibrated based on this location. The measurement benchmark points are points outside the landslide area that are laid out with reference to the centroid of the landslide and follow predetermined angle and spacing rules. These measurement benchmark points should be distributed as much as possible in the stable area around the landslide body to facilitate monitoring from these points and provide a stable reference for the layout of the identification points. In addition, the laid benchmark points should be avoided as much as possible in areas where large deformations may occur, so as not to be affected by the landslide itself, thereby affecting the accuracy of monitoring.

[0054] Based on the measurement benchmark points determined in the three-dimensional space model, these measurement benchmark points are mapped to the target landslide, and a first identification point is arranged at the corresponding position of each measurement benchmark point.

[0055] The first identification point is the reference point outside the landslide boundary. Its main function is to provide a stable reference for subsequent landslide deformation monitoring. These identification points will be part of the monitoring network. The first identification points are usually located in relatively stable areas with small deformations. Therefore, they can effectively serve as a zero point or reference point for coordinate calibration and error correction. In practical applications, the first identification point can be a simple regular identification block with central symmetry and lines of different widths on one side. It has a rough bonding surface, and the bonding surface can be consolidated with the rock and soil body by concrete, glue, etc. and maintain long-term stability. The first identification point is also generally marked with a numbered mark. By setting different lines and numbered marks, the registration efficiency and registration success rate of subsequent images can be increased, and it can also be applied to imaging devices with different resolutions.

[0056] In some optional embodiments, the acquisition of panoramic images can also be combined with passive multi-spectral optical imaging and active laser, LiDAR, InSAR and other technologies. Through the joint use of active and passive sensors, more comprehensive and accurate three-dimensional image data can be provided. Active optical imaging technology provides high-resolution visual information, while passive technologies such as LiDAR and InSAR can accurately obtain elevation data and surface deformation information under various environmental conditions, especially at night or in bad weather. Through this combination, the details and deformation of the landslide surface can be fully captured, providing more accurate and reliable three-dimensional data support for landslide monitoring.

[0057] In one embodiment, the step of arranging a plurality of second identification points within the landslide boundary of the target landslide includes: Acquiring a panoramic image of the target landslide; Based on the panoramic image, construct a three-dimensional spatial model of the target landslide; Based on the landslide evolution mechanism theory, the deformation velocity of each point of the target landslide is deduced in the three-dimensional space model; Based on the deformation speed of each of the points, the target landslide is divided into a plurality of sub-areas; According to the deformation speed of the points corresponding to each of the sub-areas, a plurality of deformation monitoring points are respectively calibrated for each of the sub-areas in the three-dimensional space model; With reference to the plurality of deformation monitoring points in the three-dimensional space model, a plurality of the second identification points are arranged within the landslide boundary of the target landslide.

[0058] In this embodiment, the panoramic image refers to a wide-angle view stitched together from high-resolution images taken from multiple angles and fields of view, which can fully display the topographic features of the landslide area. The purpose of obtaining a panoramic image is to provide sufficient image data for the subsequent construction of a three-dimensional space model. In practical applications, the panoramic image can be taken by a drone equipped with a high-definition camera, using multiple viewing angles and appropriate overlap for image acquisition. The flight path and shooting angle of the drone need to be designed according to the specific morphology and terrain conditions of the landslide to ensure that the image coverage is comprehensive and has no blind spots. In practical applications, the number of drones can be one or more. In some cases, a drone hangar can be set up for continuous regional deformation monitoring of multiple landslide groups.

[0059] In this embodiment, by applying stereoscopic vision processing technology to panoramic images, three-dimensional spatial coordinates can be extracted from image data of multiple types (passive multi-spectral optical imaging and active laser, LiDAR, InSAR, etc.) and multiple perspectives to construct a three-dimensional spatial model of the target landslide area. The three-dimensional model not only shows the surface morphology of the landslide, but also accurately reflects the changes in terrain, providing basic data for subsequent landslide monitoring.

[0060] The landslide evolution mechanism theory usually simulates the deformation process of the landslide body based on the geological causes of the landslide, climate influence, seismic activity and other factors. Through these theories, the deformation speed of points in different regions of the landslide body can be deduced and the future evolution trend of the landslide can be predicted. Therefore, in the three-dimensional space model, the landslide evolution mechanism theory can be used to deduce the deformation speed of each point of the target landslide.

[0061] According to the deduced deformation speed, the landslide area is divided into multiple sub-areas. The deformation speed of each sub-area is relatively uniform, which can effectively reflect the deformation conditions in the area. Dividing the sub-areas helps to accurately locate areas with more significant deformation and provide guidance for the subsequent deployment of monitoring points. By dividing the target landslide into multiple sub-areas, the deployment density of deformation monitoring points can be adjusted according to the deformation speed of points in different sub-areas. More deformation monitoring points are deployed in areas with faster deformation speeds to improve monitoring accuracy; in areas with slower deformation speeds, the number of deformation monitoring points can be appropriately reduced to optimize resource allocation.

[0062] Based on the deformation monitoring points determined in the three-dimensional space model, these deformation monitoring points are mapped to the target landslide, and a second identification point is arranged at the corresponding position of each deformation monitoring point. It should be noted that in this application, when arranging the second identification point, for the position points with loose soil as the base, a structure that can be inserted into the soil at the bottom is used for throwing and insertion. For the position points with a harder overall structure, concrete, glue, etc. can also be used to consolidate the bonding surface with the rock and soil and maintain long-term stability.

[0063] In some embodiments, the layout of the second identification points is also closely related to the area and deformation accuracy requirements of the target landslide. According to the characteristics of the landslide, including factors such as the area of ​​the landslide and the deformation partition, reasonable layout is carried out in combination with the actual situation of the landslide. For larger landslide areas, it is necessary to increase the density of identification points in areas with more significant deformation or higher risks to improve monitoring accuracy. In areas with slower or more stable deformation, the density of identification points can be appropriately reduced to optimize resource allocation. Through this targeted layout, more accurate and comprehensive data support can be ensured in landslide monitoring.

[0064] In one embodiment, determining the coordinate change amount of the three-dimensional coordinate of each second identification point in the two adjacent acquisition cycles according to the three-dimensional coordinates of a plurality of second identification points in each acquisition cycle in the two adjacent acquisition cycles respectively includes: Based on the three-dimensional coordinates of each of the second identification points in the two adjacent acquisition cycles, respectively determine the three-dimensional coordinates of the second identification point corresponding to each number identification in the two adjacent acquisition cycles; According to the three-dimensional coordinates of the second identification point corresponding to each numbering identification in the two adjacent acquisition cycles, the coordinate change amount of the second identification point corresponding to each numbering identification in the two adjacent acquisition cycles is determined.

[0065] In this embodiment, by comparing the three-dimensional coordinates of each second identification point in two adjacent acquisition cycles, the coordinate change of each identification point between the two cycles is calculated. Specifically, first, according to the serial number, the three-dimensional coordinate data of each second identification point in the first acquisition cycle and the second acquisition cycle are extracted. Then, the displacement change of each identification point in the X, Y, and Z directions is calculated. Next, calculate the total displacement of each second marker point using the three-dimensional distance formula ,in, is the total displacement of the second identification point, thereby obtaining the total deformation value of the second identification point. In this way, the deformation of the landslide surface can be comprehensively evaluated. Further, the coordinate changes of all identification points are summarized to generate the deformation field of the landslide, revealing which areas have more significant landslide deformation and which areas are relatively stable. These deformation data help identify high-risk areas and provide data support for real-time monitoring, risk assessment and early warning of landslides.

[0066] In one embodiment, determining the surface deformation result of the target landslide according to the coordinate change of each second identification point includes: Determine the surface deformation result of the second identification point corresponding to each number identification according to the coordinate change amount of the second identification point corresponding to each number identification within the two adjacent acquisition cycles; According to the surface deformation result of the second identification point corresponding to each serial identification, the surface deformation result of the target landslide is obtained by multi-point interpolation calculation.

[0067] In this embodiment, the coordinate change amount of each second identification point reflects the degree of deformation of the point on the landslide surface. According to the change amount of each identification point, the deformation result of the point on the landslide surface can be calculated. In this application, the deformation result can be the displacement of the point. For each second identification point, the surface deformation is determined by the coordinate change amount. For example, the displacement calculated based on the coordinate change amount can reflect the horizontal or vertical movement of the second identification point between two cycles. If a point position has undergone a significant displacement, it indicates that the landslide deformation in the area is more severe, otherwise it indicates that the area is stable or the deformation is small.

[0068] After the deformation result of each second identification point is obtained, the surface deformation result of the entire target landslide can be obtained through multi-point interpolation calculation.

[0069] This embodiment determines the surface deformation result of each second identification point by calculating the coordinate change of the point, and obtains the surface deformation result of the entire landslide area by multi-point interpolation calculation. This process can fully reflect the deformation of the landslide and help monitor and evaluate the stability of the landslide.

[0070] In some optional embodiments, both the first identification point and the second identification point can adopt a central symmetrical structure, and lines from thick to thin are designed, and each identification point is equipped with a unique number. These features help to improve the efficiency of image recognition and the success rate of registration. The central symmetrical design enables the identification points to be stably identified at different angles and viewing angles, reducing the difficulty of registration caused by image rotation or deformation. The design of lines from thick to thin enhances the recognizability of the identification points, especially in low-resolution images, and can also maintain good clarity to ensure that it can adapt to imaging devices with different resolutions. In addition, the number identification provides a unique identifier for each identification point, further improving the accuracy of automatic identification and data processing. These design features not only ensure the stability of the identification points under various conditions, but also greatly improve the accuracy and success rate of image registration, providing an accurate data basis for subsequent landslide deformation monitoring and analysis.

[0071] In one embodiment, before determining the three-dimensional coordinates of the second identification point corresponding to each serial identification in the two adjacent acquisition cycles, the method further includes: Performing image segmentation on the collected multiple images of the target landslide to obtain multiple first images including the first identification points and multiple second images including the second identification points; Based on the first image and the second image, a deep learning algorithm is used to train an image recognition model to identify the first identification point and the second identification point through the image recognition model.

[0072] In this embodiment, first, image segmentation processing is performed on multiple images collected for the target landslide. The purpose of image segmentation is to divide different areas in the image, especially to identify the part containing the first identification point and the second identification point. The segmented images are divided into two categories: one is the "first image" containing the first identification point, and the other is the "second image" containing the second identification point. This classification process can help the subsequent identification point positioning work, making the recognition process more accurate and efficient.

[0073] Next, using the segmented first image and the second image, a deep learning algorithm is used to train an image recognition model. The task of this image recognition model is to automatically identify the identification points in the image, especially the positions of the first identification point and the second identification point. Through the training process, the model will learn the position, shape and distribution of the feature points in the image, so that it can accurately extract the identification points from the image.

[0074] After the deep learning model is trained, it can be used to automatically process new images. Through this model, the positions of the first and second marker points in the image can be identified, and by calibrating their specific coordinates, they can be mapped to the three-dimensional spatial model of the landslide, thereby reducing the need for manual labeling and improving the efficiency and accuracy of marker point recognition.

[0075] In one embodiment, obtaining the three-dimensional coordinates of each point in the target landslide based on the collected multiple images includes: Performing pixel registration and pixel stitching on the multiple images collected to obtain a pixel-level composite image; Establishing a three-dimensional coordinate system based on any one of the first identification points in the pixel-level synthetic image; The three-dimensional coordinates of each point in the target landslide are determined according to the three-dimensional coordinate system.

[0076] In this embodiment, pixel registration and pixel splicing are first performed on the collected multiple images to obtain a pixel-level composite image. The process of pixel registration aligns the images collected from different perspectives or time points to ensure their precise matching in space. Splicing seamlessly synthesizes these images to form a complete image covering the entire landslide area, ensuring the integrity and consistency of the image data. Next, a first identification point in the composite image is selected as a reference point to establish a three-dimensional coordinate system. The reference point is usually located in a stable area and can be used as a reference point to define the coordinate system after accurate positioning. Finally, according to the established three-dimensional coordinate system, the three-dimensional coordinates of each point in the target landslide are calculated through the corresponding relationship between the pixel position in the image and the actual three-dimensional coordinates. In this way, each point in the image can be accurately mapped to the three-dimensional space, providing accurate data support for landslide deformation monitoring and analysis.

[0077] In one embodiment, after determining the surface deformation result of the target landslide, the method further includes: Obtaining surface deformation results corresponding to the target landslide in multiple groups of acquisition cycles, wherein a group of acquisition cycles includes two adjacent acquisition cycles; The plurality of surface deformation results are fused to obtain a continuous deformation field of the target landslide within the plurality of acquisition cycles.

[0078] This embodiment describes how to obtain the continuous deformation field of the target landslide in multiple acquisition cycles by further processing the data of multiple acquisition cycles after determining the surface deformation result of the target landslide. The specific steps are as follows: A set of acquisition cycles consists of two adjacent acquisition cycles, and each acquisition cycle obtains the surface deformation data of the target landslide by monitoring the surface deformation of the landslide. The data of multiple acquisition cycles provide deformation results at multiple time points for the subsequent continuous deformation field construction.

[0079] In this embodiment, in each group of acquisition cycles, the surface deformation results of the target landslide in each acquisition cycle are obtained by image processing, three-dimensional coordinate calculation and interpolation. These deformation results reflect the displacement and deformation of the landslide in each time period, and are usually presented in the form of displacement field or deformation field.

[0080] In order to obtain the continuous deformation field of the target landslide in multiple acquisition cycles, the surface deformation results in each acquisition cycle need to be fused. The fusion method can be weighted average, interpolation or least squares fitting, etc. The purpose is to merge the deformation results in different time periods into a continuous deformation model. Through fusion, not only the overall deformation of the target landslide can be obtained, but also the errors or data inconsistencies that may exist between different cycles can be eliminated. The fused results form a smooth and continuous deformation field, showing the deformation trend of the landslide at multiple time points. This continuous deformation field can reveal the evolution process of landslide deformation and provide more comprehensive and detailed monitoring data.

[0081] This embodiment fuses the surface deformation results of the target landslide in multiple acquisition cycles to obtain a continuous deformation field that can intuitively display the deformation dynamics of the landslide. This deformation field reflects the overall displacement of the landslide in multiple time periods, providing key data support for landslide risk assessment and early warning.

[0082] In one embodiment, after determining the surface deformation result of the target landslide, the method further includes: Obtaining correction values ​​of deformation results of the target landslide collected by a plurality of total station measuring points pre-arranged within the landslide boundary of the target landslide; The surface deformation result of the target landslide is corrected according to the multiple correction values.

[0083] In this embodiment, the total station measuring point is a reference point pre-arranged within the landslide boundary, usually located in a stable area within the landslide body, and can provide high-precision displacement data. The total station can monitor the position changes of these measuring points in real time and accurately measure the deformation of each measuring point during the landslide process. Since the total station measuring points usually have a high degree of accuracy, the data they provide can be used as a "reference standard" for landslide deformation monitoring. After the surface deformation results of the target landslide are calculated, the deformation correction value of each measuring point can be obtained by comparing the actual deformation data of each measuring point of the landslide (obtained by the total station) with the preliminary calculated surface deformation results. The correction value is usually calculated based on the difference between the actual displacement measured by the total station and the displacement predicted by the model. After obtaining the deformation correction values ​​of the total station measuring points, these correction values ​​can be applied to the preliminary calculated surface deformation results of the target landslide. Specifically, the deformation data of the entire landslide area can be adjusted according to the correction value of each measuring point. In practical applications, the correction value can be proportionally distributed to the entire landslide area to correct the error in the preliminary calculation and improve the accuracy of the final deformation result.

[0084] The present application provides a method for monitoring deformation of a landslide surface, the method comprising: determining a landslide boundary of a target landslide to be monitored; arranging a plurality of first identification points outside the landslide boundary of the target landslide, and arranging a plurality of second identification points within the landslide boundary of the target landslide, wherein each of the second identification points is provided with a number identification; based on the plurality of first identification points, setting a plurality of patrol paths, collecting a plurality of images of the target landslide by means of a drone according to the plurality of patrol paths, and obtaining a three-dimensional coordinate of each point in the target landslide based on the collected plurality of images; determining the three-dimensional coordinates of a plurality of second identification points in each of two adjacent acquisition cycles according to the three-dimensional coordinates of each point in the target landslide; determining the coordinate change amount of the three-dimensional coordinate of each of the second identification points in the two adjacent acquisition cycles according to the three-dimensional coordinates of the plurality of second identification points in each of the two adjacent acquisition cycles; determining the surface deformation result of the target landslide according to the coordinate change amount of each of the second identification points. The present application first determines the boundary of the target landslide, and arranges a first identification point outside the boundary of the landslide and a second identification point with a number inside the boundary; then, multiple drone patrol paths are set based on the first identification point, multiple images of the target landslide are collected and the three-dimensional coordinates of each point are calculated; then, the three-dimensional coordinate changes of the second identification point in two adjacent collection cycles are compared, and the coordinate change amount is calculated; finally, the surface deformation result of the target landslide is determined according to the coordinate change amount, and the first identification point is arranged outside the boundary of the landslide to achieve high-precision positioning of each point of the target landslide, and the drone patrol path is set according to the first identification point to monitor the second identification point within the boundary of the landslide, so as to achieve all-round monitoring of the deformation of the target landslide.

[0085] Based on the same inventive concept, the second aspect of the present application provides a landslide surface deformation monitoring system, such as Figure 2 As shown, the system comprises: A determination module 201 is used to determine the landslide boundary of the target landslide to be monitored; A layout module 202 is used to layout a plurality of first identification points outside the landslide boundary of the target landslide, and to layout a plurality of second identification points inside the landslide boundary of the target landslide, wherein each of the second identification points is provided with a number identification; A setting module 203 is used to set a plurality of patrol paths based on the plurality of first identification points, collect a plurality of images of the target landslide by using a drone according to the plurality of patrol paths, and obtain the three-dimensional coordinates of each point in the target landslide based on the collected plurality of images; The first determination module 204 is used to determine the three-dimensional coordinates of a plurality of the second identification points in each acquisition cycle in two adjacent acquisition cycles according to the three-dimensional coordinates of each point in the target landslide, wherein one acquisition cycle is defined as one image acquisition completed by the UAV along all patrol paths; A second determination module 205 is used to determine the coordinate change amount of the three-dimensional coordinate of each second identification point in the two adjacent acquisition cycles according to the three-dimensional coordinates of a plurality of second identification points in each acquisition cycle in the two adjacent acquisition cycles; The third determination module 206 is used to determine the surface deformation result of the target landslide according to the coordinate change of each of the second identification points.

[0086] Optionally, the arranging of a plurality of first identification points outside the landslide boundary of the target landslide, the arranging module 202 includes: A first acquisition submodule is used to acquire a panoramic image of the target landslide; A first construction submodule is used to construct a three-dimensional spatial model of the target landslide based on the panoramic image; A first determination submodule is used to determine the landslide centroid of the target landslide in the three-dimensional space model by taking the landslide boundary of the three-dimensional space model as a reference; A first calibration submodule is used to calibrate a plurality of measurement reference points outside the landslide boundary in the three-dimensional space model at preset interval angles, using the landslide centroid as a reference point; The first layout submodule is used to layout a plurality of the first identification points outside the landslide boundary of the target landslide with reference to the plurality of measurement reference points in the three-dimensional space model.

[0087] Optionally, the arranging of a plurality of second identification points within the landslide boundary of the target landslide, the arranging module 202 comprises: A second acquisition submodule is used to acquire a panoramic image of the target landslide; A second construction submodule is used to construct a three-dimensional spatial model of the target landslide based on the panoramic image; A deduction submodule, for deducing the deformation velocity of each point of the target landslide in the three-dimensional space model based on the evolution mechanism theory of the landslide; A region division submodule, used for dividing the target landslide into a plurality of sub-regions based on the deformation velocity of each of the points; A second calibration submodule is used to calibrate a plurality of deformation monitoring points for each of the sub-areas in the three-dimensional space model according to the deformation speed of the points corresponding to each of the sub-areas; The second layout submodule is used to layout a plurality of the second identification points within the landslide boundary of the target landslide with reference to the plurality of the deformation monitoring points in the three-dimensional space model.

[0088] Optionally, the coordinate change amount of the three-dimensional coordinate of each second identification point in the two adjacent acquisition cycles is determined respectively according to the three-dimensional coordinates of a plurality of second identification points in each acquisition cycle in the two adjacent acquisition cycles, and the second determination module 205 includes: A second determination submodule, configured to determine the three-dimensional coordinates of the second identification point corresponding to each numbering identification in the two adjacent acquisition periods based on the three-dimensional coordinates of each of the second identification points in the two adjacent acquisition periods; The third determination submodule is used to determine the coordinate change of the second identification point corresponding to each numbering identification within the two adjacent acquisition cycles according to the three-dimensional coordinates of the second identification point corresponding to each numbering identification within the two adjacent acquisition cycles.

[0089] Optionally, the third determining module 206 includes: A fourth determination submodule, configured to determine the surface deformation result of the second identification point corresponding to each number identification according to the coordinate change amount of the second identification point corresponding to each number identification within the two adjacent acquisition cycles; The calculation submodule is used to obtain the surface deformation result of the target landslide through multi-point interpolation calculation according to the surface deformation result of the second identification point corresponding to each serial identification.

[0090] Optionally, the system further comprises: An image segmentation submodule, used for performing image segmentation on the multiple images of the target landslide collected to obtain multiple first images containing the first identification points and multiple second images containing the second identification points; An image recognition submodule is used to train an image recognition model using a deep learning algorithm based on the first image and the second image, so as to identify the first identification point and the second identification point through the image recognition model.

[0091] Optionally, the three-dimensional coordinates of each point in the target landslide are obtained based on the collected multiple images, and the setting module 203 includes: A registration and stitching submodule, used for performing pixel registration and pixel stitching on the multiple images collected to obtain a pixel-level composite image; An establishing submodule, used for establishing a three-dimensional coordinate system based on any one of the first identification points in the pixel-level synthetic image as a reference; The fifth determination submodule is used to determine the three-dimensional coordinates of each point in the target landslide according to the three-dimensional coordinate system.

[0092] Optionally, the system further comprises: A third acquisition submodule is used to acquire surface deformation results corresponding to the target landslide in multiple groups of acquisition cycles, wherein a group of acquisition cycles includes two adjacent acquisition cycles; The fusion submodule is used to fuse the multiple surface deformation results to obtain the continuous deformation field of the target landslide within the multiple groups of acquisition cycles.

[0093] Optionally, the system further comprises: A fourth acquisition submodule is used to acquire a correction value of a deformation result of the target landslide collected by each of a plurality of total station measuring points pre-arranged within the landslide boundary of the target landslide; The correction submodule is used to correct the surface deformation result of the target landslide according to the multiple correction values.

[0094] Based on the same inventive concept, the third aspect of the present application provides a Figure 3 The electronic device 100 shown includes a memory 110, a processor 120 and a computer program stored in the memory 110, and the processor 120 executes the computer program to implement the landslide surface deformation monitoring method as described in the first aspect of the present application.

[0095] In a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the landslide surface deformation monitoring method as described in the first aspect of the present application is implemented.

[0096] Each embodiment in this specification focuses on the differences from other embodiments. The same and similar parts between the embodiments may be referred to each other.

[0097] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the embodiments of the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0098] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0099] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0101] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present application.

[0102] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "including a..." do not exclude the presence of other identical elements in the process, method, article or terminal device including the elements.

[0103] The above is a detailed introduction to a landslide surface deformation monitoring method, system, equipment and medium provided. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for monitoring landslide surface deformation, characterized in that: The method comprises: Determine the landslide boundary of the target landslide to be monitored; Arranging a plurality of first identification points outside the landslide boundary of the target landslide, and arranging a plurality of second identification points inside the landslide boundary of the target landslide, wherein each of the second identification points is provided with a number identification; Based on the plurality of first identification points, a plurality of patrol paths are set, and a plurality of images of the target landslide are respectively collected by the drone according to the plurality of patrol paths, and based on the collected plurality of images, the three-dimensional coordinates of each point in the target landslide are obtained; According to the three-dimensional coordinates of each point in the target landslide, the three-dimensional coordinates of a plurality of the second identification points in each acquisition cycle in two adjacent acquisition cycles are determined respectively, wherein one acquisition cycle is when the drone completes one image acquisition along all patrol paths; According to the three-dimensional coordinates of a plurality of the second identification points in each of the two adjacent acquisition cycles, respectively determining a coordinate change amount of the three-dimensional coordinate of each of the second identification points in the two adjacent acquisition cycles; The surface deformation result of the target landslide is determined according to the coordinate change of each of the second identification points.

2. The method for monitoring landslide surface deformation according to claim 1, characterized in that: The step of arranging a plurality of first identification points outside the landslide boundary of the target landslide comprises: Acquiring a panoramic image of the target landslide; Based on the panoramic image, construct a three-dimensional spatial model of the target landslide; Taking the landslide boundary of the three-dimensional space model as a reference, determining the landslide centroid of the target landslide in the three-dimensional space model; Using the centroid of the landslide as a reference point, calibrating a plurality of measurement reference points outside the landslide boundary in the three-dimensional space model at preset interval angles; With reference to the plurality of measurement reference points in the three-dimensional space model, a plurality of the first identification points are arranged outside the landslide boundary of the target landslide.

3. The method for monitoring landslide surface deformation according to claim 1, characterized in that: The step of arranging a plurality of second identification points within the landslide boundary of the target landslide comprises: Acquiring a panoramic image of the target landslide; Based on the panoramic image, construct a three-dimensional spatial model of the target landslide; Based on the landslide evolution mechanism theory, the deformation velocity of each point of the target landslide is deduced in the three-dimensional space model; Based on the deformation speed of each of the points, the target landslide is divided into a plurality of sub-areas; According to the deformation speed of the points corresponding to each of the sub-areas, a plurality of deformation monitoring points are respectively calibrated for each of the sub-areas in the three-dimensional space model; With reference to the plurality of deformation monitoring points in the three-dimensional space model, a plurality of the second identification points are arranged within the landslide boundary of the target landslide.

4. The method for monitoring landslide surface deformation according to claim 1, characterized in that: The step of determining the coordinate change amount of the three-dimensional coordinate of each second identification point in the two adjacent acquisition cycles according to the three-dimensional coordinates of the plurality of second identification points in each acquisition cycle in the two adjacent acquisition cycles comprises: Based on the three-dimensional coordinates of each of the second identification points in the two adjacent acquisition cycles, respectively determine the three-dimensional coordinates of the second identification point corresponding to each number identification in the two adjacent acquisition cycles; According to the three-dimensional coordinates of the second identification point corresponding to each numbering identification in the two adjacent acquisition cycles, the coordinate change amount of the second identification point corresponding to each numbering identification in the two adjacent acquisition cycles is determined.

5. The method for monitoring landslide surface deformation according to claim 4, characterized in that: Determining the surface deformation result of the target landslide according to the coordinate change of each second identification point includes: Determine the surface deformation result of the second identification point corresponding to each number identification according to the coordinate change amount of the second identification point corresponding to each number identification within the two adjacent acquisition cycles; According to the surface deformation result of the second identification point corresponding to each serial identification, the surface deformation result of the target landslide is obtained by multi-point interpolation calculation.

6. The method for monitoring landslide surface deformation according to claim 4, characterized in that: Before determining the three-dimensional coordinates of the second identification point corresponding to each number identification in the two adjacent acquisition cycles, the method further includes: Performing image segmentation on the collected multiple images of the target landslide to obtain multiple first images including the first identification points and multiple second images including the second identification points; Based on the first image and the second image, a deep learning algorithm is used to train an image recognition model to identify the first identification point and the second identification point through the image recognition model.

7. The method for monitoring landslide surface deformation according to claim 1, characterized in that: The step of obtaining the three-dimensional coordinates of each point in the target landslide based on the collected multiple images includes: Performing pixel registration and pixel stitching on the multiple images collected to obtain a pixel-level composite image; Establishing a three-dimensional coordinate system based on any one of the first identification points in the pixel-level synthetic image; The three-dimensional coordinates of each point in the target landslide are determined according to the three-dimensional coordinate system.

8. The method for monitoring landslide surface deformation according to claim 1, characterized in that: After determining the surface deformation result of the target landslide, the method further includes: Obtaining surface deformation results corresponding to the target landslide in multiple groups of acquisition cycles, wherein a group of acquisition cycles includes two adjacent acquisition cycles; The plurality of surface deformation results are fused to obtain a continuous deformation field of the target landslide within the plurality of acquisition cycles.

9. The method for monitoring landslide surface deformation according to claim 1, characterized in that: After determining the surface deformation result of the target landslide, the method further includes: Obtaining correction values ​​of deformation results of the target landslide collected by a plurality of total station measuring points pre-arranged within the landslide boundary of the target landslide; The surface deformation result of the target landslide is corrected according to the multiple correction values.

10. A landslide surface deformation monitoring system, characterized in that: The system comprises: A determination module, used for determining the landslide boundary of the target landslide to be monitored; A layout module, used for arranging a plurality of first identification points outside the landslide boundary of the target landslide, and arranging a plurality of second identification points inside the landslide boundary of the target landslide, wherein each of the second identification points is provided with a number identification; A setting module, used to set a plurality of patrol paths based on the plurality of first identification points, collect a plurality of images of the target landslide by the drone according to the plurality of patrol paths, and obtain the three-dimensional coordinates of each point in the target landslide based on the collected plurality of images; A first determination module is used to determine the three-dimensional coordinates of a plurality of the second identification points in each acquisition cycle in two adjacent acquisition cycles according to the three-dimensional coordinates of each point in the target landslide, wherein one acquisition cycle is when the drone completes one image acquisition along all patrol paths; A second determination module, configured to determine, based on the three-dimensional coordinates of a plurality of the second identification points in each of the two adjacent acquisition cycles, a coordinate change amount of the three-dimensional coordinates of each of the second identification points in the two adjacent acquisition cycles; The third determination module is used to determine the surface deformation result of the target landslide according to the coordinate change of each of the second identification points.

11. An electronic device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the landslide surface deformation monitoring method according to any one of claims 1 to 9.

12. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the landslide surface deformation monitoring method according to any one of claims 1 to 9 is implemented.

Images