A method and device for deformation monitoring using a drone

Abstract

The present application relates to a kind of method and equipment for deformation monitoring using unmanned aerial vehicle, belong to surveying and mapping, photogrammetry and water conservancy monitoring technical field.The same detection range is used unmanned aerial vehicle to complete the aerial data acquisition of two time, and the photograph with the degree of overlap is collected;Image or video frame after image acquisition, extract the unmanned aerial vehicle attitude and GNSS space coordinates when image or image frame is photographed;Restore the spatial attitude of photograph, find the homonym of photograph, restore the relative position between photograph, calculate the digital elevation model of ground surface of two time;Compare the digital elevation model of two time, first alignment, extract the same space range data of DEM of two time, then difference calculation is carried out to the aligned DEM of two time, and if elevation difference exceeds threshold value, it indicates that there is collapse, and the earthwork of collapse area is calculated according to DEM;The collapse coordinate position is positioned to GIS system and visualized display.The present application can quickly spatial positioning, show defect, identify whether there is collapse.

Description

Technical Field

[0001] This invention relates to a method and equipment for deformation monitoring using unmanned aerial vehicles (UAVs), belonging to the fields of surveying, photogrammetry, and water conservancy monitoring technology. Background Technology

[0002] The Yellow River has engineering dams to prevent the river from eroding the riverbed. These dams are at risk of collapsing due to water. To eliminate this risk, regular inspections of the dams are necessary. However, manual inspections are inefficient, have limited visibility, and there are hidden dangers in riverbank patrols.

[0003] Existing technologies include those that utilize remote sensing data to accurately extract the distribution of elevation changes in the study area through high-precision DEM data, effectively reflecting the topographic changes before the earthquake, such as patent CN 110007336 A, but these technologies have relative limitations and shortcomings.

[0004] Drones offer a wide field of view from the air, allowing for multi-angle inspections of targets. Their low flight altitude and high camera resolution enable them to identify small objects. Aerial photographs taken by drones can be used to create 3D models, and comparing these models at different times during inspections of the same location can help identify potential collapses in engineering projects like dams. Summary of the Invention

[0005] The purpose of this invention is to overcome the above-mentioned shortcomings and provide a method for deformation monitoring using drones, which can quickly locate in space, display defects, and identify whether there is a collapse.

[0006] The technical solution adopted in this invention is as follows: A method for deformation monitoring using drones includes the following steps: S1. Use drones to collect aerial data at two different times within the same detection range, and collect photos with overlap. S2. Obtain image frames after image or video frame extraction, and extract the UAV attitude and GNSS spatial coordinates at the time the image or image frame was captured; S3. Restore the spatial orientation of the photos, find the corresponding points in the photos, restore the relative positions between the photos, and calculate the digital elevation model of the ground surface at the two time points. S4. Compare the digital elevation models at two different times. First, align them and extract the spatial range data of the DEMs at the two times. Then, calculate the difference between the aligned DEMs at the two times. If the elevation difference exceeds the threshold, it indicates a collapse. Calculate the earthwork volume of the collapse area based on the DEM. S5. Locate the collapse coordinates in the GIS system and display them visually.

[0007] The surface digital elevation model calculation in step S3 of the above method uses a bundle adjustment model, which is obtained through forward intersection in the following matrix form: , , , , Image point coordinates, For the coordinates of ground control points, For self-calibration parameters, Corrections for the station coordinate observations; For the increment of unknown coordinates of the encryption point, c is the increment of unknown elements in the exterior orientation of the photograph, and c is the self-calibration parameter. The increment of the unknown component of the eccentricity between the antenna phase center and the shooting center. This is the drift error correction value; , , In response to The coefficient; yes The coefficient; It is the identity matrix; It is a matrix with constant coefficients; This is the observation weight matrix. (The variables are all matrices, and commas are used to separate columns).

[0008] The threshold for determining collapse in step S5 is determined based on the measurement error of this method and the size of the collapse to be detected. The volume of earthwork in the collapse area can be calculated by summing the products of the area of ​​each cell in the changed section and the elevation difference.

[0009] A device for deformation monitoring using unmanned aerial vehicles (UAVs) includes an airborne acquisition device, an image preprocessing module, a digital elevation model (DEM) calculation module, a two-time-time DEM comparison module, a collapse zone analysis and calculation module, and a visualization module. Airborne data acquisition equipment is responsible for collecting aerial data from the same detection range at two different times using drones, and collecting photos with overlapping characteristics. The image preprocessing module acquires image frames after image or video frame extraction, and extracts the UAV attitude and GNSS spatial coordinates at the time the image or image frame was captured. The surface digital elevation model calculation module is used to obtain the DEM and provide calculation data for comparison. The two-time digital elevation model comparison module converts the surface digital elevation model into the same geospatial coordinate system, aligns them, extracts spatial range data of the two-time DEMs, and calculates the difference between the aligned two-time DEMs. The collapse zone analysis and calculation module marks areas with elevation differences exceeding a threshold as collapsed and calculates the earthwork volume of the collapsed area based on the DEM. The visualization module locates the collapse coordinates and displays them visually in the GIS system.

[0010] The beneficial effects of this invention are as follows: (1) Low monitoring cost, pure optical solution, no need for lidar; (2) Eliminate manual inspections and take advantage of air power to reach places that are not easily accessible by humans; (3) It is easy to obtain the volume of collapsed earthwork and easy to embed into the GIS system for decision analysis (spatial positioning, navigation, etc.). (4) The data scale detected by the method of the present invention is larger and can detect more detailed changes. Compared with the difficulty of satellite orbit scheduling, UAVs can be scheduled for detection at any time. Attached Figure Description

[0011] Figure 1 Flowchart of the method of this invention; Figure 2 The simulation digital elevation model comparison diagrams in the method of this invention are shown as follows: a. Scene 1, b. Scene 2, c. Scene 1 DEM, d. Scene 2 DEM, e. Overlay change of the two scene DEMs; Figure 3 shows the data collected by the UAV. a is an overview of the overall aerial survey, and b is the observation map of the engineering dam. Figure 4 This is a schematic diagram illustrating the principle of the beam method. Detailed Implementation

[0012] The present invention will be further described below with reference to specific embodiments.

[0013] Example 1: A method for deformation monitoring using a drone, comprising the following steps: Interior orientation elements: parameters relating the position between the camera center and the image; exterior orientation elements: parameters determining the spatial position and orientation of the camera beam at the moment of capture.

[0014] S1. Use a drone to collect aerial data at two different times within the same detection range, capturing photos with overlap: Prepare the drone, calibration board, computer, and programming software; calibrate the camera, acquire interior orientation elements, and eliminate camera manufacturing errors; collect data at two time points: aerial data at time t1 and aerial data at time t2.

[0015] S2. Acquire image frames after image or video frame extraction, and extract the UAV attitude and GNSS spatial coordinates at the time the image or image frame was captured: If the video is saved, it is necessary to extract frames from the aerial video and extract the drone's attitude and GNSS spatial coordinates at each frame. Generally, the drone's captioning function will record this information. Errors caused by camera manufacturing should also be removed from the image.

[0016] S3. Restore the spatial orientation of the photographs, find corresponding points in the photographs, restore the relative positions between the photographs, and calculate the digital elevation model of the Earth's surface at the two time points: The aforementioned digital elevation model for the Earth's surface is calculated using a bundle adjustment model, obtained through forward intersection in the following matrix form: , , , , V X Image point coordinates, For the coordinates of ground control points, For self-calibration parameters, Corrections for the station coordinate observations; For the increment of unknown coordinates of the encryption point, c is the increment of unknown elements in the exterior orientation of the photograph, and c is the self-calibration parameter. The increment of the unknown component of the eccentricity between the antenna phase center and the shooting center. This is the drift error correction value; , , In response to The coefficient; yes The coefficient; It is the identity matrix; It is a matrix with constant coefficients; This is the observation weight matrix.

[0017] Steps for calculating a digital elevation model: a) Determination of the approximate coordinates of exterior orientation elements and ground points in each image, which includes GNSS data and attitude data collected by the UAV.

[0018] b. Starting from the image coordinates of the control points and the points to be determined on each image, list the error equations according to the collinearity condition equations for each photographic ray.

[0019] , , , , , , , , , , , , , , For interior orientation elements; For exterior azimuth elements; () represents the image point coordinates. The coordinates of the ground point, To correct the number, Z is the auxiliary coordinate in space; a 11 a 12 a 13 …a 26 The partial derivatives of each variable, Indicates parameters to be determined. This represents the correction number for the parameter to be determined.

[0020] c. Point-by-point transformation: Establish the transformation equation, and first obtain the exterior orientation element t of each image using the cyclic block solution method: , , , , , Represents a constant term. If the difference is less than the preset value, the iterative calculation ends.

[0021] d. After obtaining the exterior orientation elements, the ground coordinates of the point to be determined are obtained through spatial forward intersection. For common intersection points of adjacent images, their average value should be taken as the final result. The calculation process of determining the object space coordinates of this point from the exterior and interior orientation elements of the left and right images of the stereo image pair and the image coordinate values ​​of corresponding points is called spatial forward intersection.

[0022] e. Dense point clouds are generated through dense matching algorithms, and digital elevation models are obtained through point cloud filtering and interpolation.

[0023] S4. Compare the digital elevation models at two different times. First, align them and extract the spatial range data of the DEMs at the two times. Then, calculate the difference between the aligned DEMs at the two times. If the elevation difference exceeds the threshold, it indicates a collapse. Calculate the earthwork volume of the collapse area based on the DEM.

[0024] Comparison of digital elevation models at two different times: a) Alignment: In actual measurement results, the spatial extent of the DEM obtained at different times for the same location is different. Extract the spatial extent of the DEM at two times and intersect them to obtain a common extent. Use this extent to crop the DEM at the two times.

[0025] b. Compare and calculate the difference between the aligned DEMs at two different times. If there is a collapsed section, calculate the earthwork volume of the collapsed area based on the DEM.

[0026] S5. Locate the collapse coordinates and visualize them in the GIS system: Analyzing the difference data, if the elevation difference exceeds a threshold, it indicates that the dam has collapsed. The coordinates can be located and displayed visually in a GIS system. The volume of earthwork caused by the collapse is also displayed.

[0027] Example 2: A device for deformation monitoring using a drone, comprising an airborne acquisition device, an image preprocessing module, a surface digital elevation model calculation module, a two-time-time digital elevation model comparison module, a collapse zone analysis and calculation module, and a visualization module. Airborne data acquisition equipment is responsible for collecting aerial data from the same detection range at two different times using drones, and collecting photos with overlapping characteristics. The image preprocessing module acquires image frames after image or video frame extraction, and extracts the UAV attitude and GNSS spatial coordinates at the time the image or image frame was captured. The surface digital elevation model calculation module is used to obtain the DEM and provide calculation data for comparison. The two-time digital elevation model comparison module converts the surface digital elevation model into the same geospatial coordinate system, aligns them, extracts spatial range data of the two-time DEMs, and calculates the difference between the aligned two-time DEMs. The collapse zone analysis and calculation module marks areas with elevation differences exceeding a threshold as collapsed and calculates the earthwork volume of the collapsed area based on the DEM. The visualization module locates the collapse coordinates and displays them visually in the GIS system.

[0028] The above is a further description of the present invention in conjunction with specific embodiments, and the scope of protection of the present invention is not limited thereto.

Claims

1. A method for deformation monitoring using unmanned aerial vehicles (UAVs), characterized in that, The steps include the following: S1. Use drones to collect aerial data at two different times within the same detection range, and collect photos with overlap. S2. Obtain image frames after image or video frame extraction, and extract the UAV attitude and GNSS spatial coordinates at the time the image or image frame was captured; S3. Restore the spatial orientation of the photos, find the corresponding points in the photos, restore the relative positions between the photos, and calculate the digital elevation model of the ground surface at the two time points. S4. Compare the digital elevation models at two different times. First, align them and extract the spatial range data of the DEMs at the two times. Then, calculate the difference between the aligned DEMs at the two times. If the elevation difference exceeds the threshold, it indicates a collapse. Calculate the earthwork volume of the collapse area based on the DEM. S5. Locate the collapse coordinates in the GIS system and display them visually.

2. The method for deformation monitoring using a drone according to claim 1, characterized in that, The surface digital elevation model calculation described in step S3 uses a bundle adjustment model, obtained through the forward intersection of the following matrices: ， ， ， ， Image point coordinates, For the coordinates of ground control points, For self-calibration parameters, Corrections for the station coordinate observations; For the increment of unknown coordinates of the encryption point, c is the increment of unknown elements in the exterior orientation of the photograph, and c is the self-calibration parameter. The increment of the unknown component of the eccentricity between the antenna phase center and the shooting center. This is the drift error correction value; , , In response to The coefficient; yes The coefficient; It is the identity matrix; It is a matrix with constant coefficients; This is the observation weight matrix.

3. The method for deformation monitoring using a drone according to claim 1, characterized in that, In step S5, the earthwork volume of the collapsed area can be calculated by the sum of the products of the area of ​​each cell in the changed part and the elevation difference.

4. A device for deformation monitoring using unmanned aerial vehicles (UAVs), characterized in that, It includes airborne acquisition equipment, an image preprocessing module, a surface digital elevation model calculation module, a two-time digital elevation model comparison module, a collapse zone analysis and calculation module, and a visualization module. Airborne data acquisition equipment is responsible for collecting aerial data from the same detection range at two different times using drones, and collecting photos with overlapping characteristics. The image preprocessing module acquires image frames after image or video frame extraction, and extracts the UAV attitude and GNSS spatial coordinates at the time the image or image frame was captured. The surface digital elevation model calculation module is used to obtain the DEM and provide calculation data for comparison. The two-time digital elevation model comparison module converts the surface digital elevation model into the same geospatial coordinate system, aligns them, extracts spatial range data of the two-time DEMs, and calculates the difference between the aligned two-time DEMs. The collapse zone analysis and calculation module marks areas with elevation differences exceeding a threshold as collapsed and calculates the earthwork volume of the collapsed area based on the DEM. The visualization module locates the collapse coordinates and displays them visually in the GIS system.

Images