Highway facility three-dimensional displacement high-precision measurement method based on low altitude

Abstract

The invention relates to the technical field of highway facility state monitoring, in particular to a highway facility three-dimensional displacement high-precision measurement method based on low altitude, which comprises the following steps of: shooting a monitoring image of a side slope by adopting a high-definition camera carried by an unmanned aerial vehicle, analyzing a reference target in the monitoring image, judging the displacement change degree of the reference target, and calculating the displacement change degree of the reference target. When the first displacement change value of the reference target is smaller than a first preset value, analyzing the monitoring image by adopting a first preset analysis rule, and determining the displacement value of each target on each axis in the three-dimensional coordinate system; and when the first displacement change value is greater than or equal to a first preset value, a target reference point is determined firstly, then a three-dimensional coordinate system is formed according to the target reference point, the displacement change value of each target is analyzed, and the displacement value of each target is determined. By means of the mode, the problems that only data in the two-dimensional direction can be obtained through a traditional method, a plurality of clinometers need to be arranged, construction is complex, and the later maintenance cost is high are solved.

Description

Technical Field

[0001] The invention relates to the technical field of highway facility status monitoring, and in particular to a high-precision measurement method for three-dimensional displacement of highway facilities based on low altitude. Background Art

[0002] Highway facilities are an important part of road engineering and have a significant impact on road safety and reliability. Among them, slope safety is crucial to road safety. Due to the influence of geological and external weather factors, slopes are prone to sliding, collapse, spalling and soil erosion, which have an adverse impact on traffic operations and seriously endanger the safety of passing vehicles. Therefore, it is very important to monitor the safety of slopes. At present, many technical solutions for slope safety monitoring have been formed and have been widely promoted and applied.

[0003] However, due to the defects of low accuracy and incomplete coverage in the existing technology, the slope monitoring data may have errors or false alarms, which greatly increases the workload of relevant personnel and cannot guarantee accurate monitoring and timely warning of the slope. At present, regarding the monitoring of slope safety status, the patent with the patent publication number CN117288920A discloses that the stress change index and the slope point displacement index are calculated and analyzed by sensors to obtain slope information, but the index obtained by this method cannot intuitively reflect the displacement of the slope in all directions; and the patent with the patent publication number CN219103989U discloses that the horizontal and vertical bidirectional displacement of the slope is monitored by using an inclinometer and a settlement monitoring component, but the data obtained by this method is still not comprehensive enough, and can only obtain data in two-dimensional directions, and multiple inclinometers need to be set up, which is complicated to construct and has high maintenance costs in the later stage. In addition, the existing slope displacement analysis methods are all based on fixed reference targets for displacement analysis, and the status of the fixed reference targets is not analyzed. When there is a problem with the fixed reference target, the measurement accuracy of the slope displacement will be reduced, resulting in inaccurate measurement results.

[0004] Therefore, the present invention provides a high-precision measurement method for three-dimensional displacement of highway facilities based on low altitude to solve the above problems. Summary of the invention

[0005] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides an immersive virtual character interaction method based on virtual reality to solve the problem that only two-dimensional data can be obtained, multiple inclinometers need to be set up, the construction is complicated, and the subsequent maintenance cost is high.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is:

[0007] A high-precision measurement method for three-dimensional displacement of highway facilities based on low altitude, comprising:

[0008] Obtain monitoring images of various angles taken by drones at low altitudes and the corresponding initial setting information;

[0009] Analyzing the state of the reference target according to the monitoring image to determine a first displacement change value of the reference target;

[0010] When the first displacement change value is less than the first preset value, analyzing and processing the monitoring image based on the first preset analysis rule to determine the displacement value of each target;

[0011] When the first displacement change value is greater than or equal to the first preset value, an initial three-dimensional model constructed by initial setting information is obtained, and a monitoring three-dimensional model of the current slope state is constructed according to the monitoring image; the initial three-dimensional model and the monitoring three-dimensional model are compared and analyzed to determine the target reference point; the initial three-dimensional model and the monitoring three-dimensional model are feature simplified to obtain a first simplified model and a second simplified model respectively; the target reference points in the first simplified model and the second simplified model are overlapped to determine the displacement value of the slope according to the first preset analysis rule.

[0012] Preferably, a low-altitude high-precision measurement method for three-dimensional displacement of highway facilities also includes: when the first displacement change value is greater than or equal to a first preset value, performing a comparison analysis based on the initial setting information and the monitoring image to determine a second reference target; according to the second reference target, analyzing and processing the monitoring image based on the first preset analysis rule to determine the displacement value of each target.

[0013] Preferably, the analysis of the state of the reference target according to the monitoring image to determine the first displacement change value of the reference target includes: preprocessing each monitoring image, identifying the reference target and extracting features to obtain the reference target features; matching and analyzing the initial target features with the reference target features, overlapping one of the reference target features in the same direction, and calculating the offset value of the initial target feature and the reference target feature; and using the two offset values ​​as the first displacement change value.

[0014] Preferably, the analyzing and processing of the monitoring images based on the first preset analysis rule to determine the displacement value of each target includes: performing target recognition and feature extraction on each monitoring image, and constructing a three-dimensional model based on the extracted features; forming a three-dimensional coordinate system with reference target features; determining the measurement mapping amount of each measurement target on each axis according to the three-dimensional coordinate system; and analyzing the measurement mapping amount according to the initial mapping amount to determine the displacement value of each target.

[0015] Preferably, the comparative analysis of the initial three-dimensional model and the monitoring three-dimensional model to determine the target reference point includes: overlapping the initial three-dimensional model and the monitoring three-dimensional model according to the initially set direction to obtain a three-dimensional model formed by overlapping feature points; performing reference point analysis of the overlapping feature points in the three-dimensional coordinate system to obtain overlapping feature matching pairs; identifying and analyzing the object areas corresponding to the overlapping feature matching pairs to determine target overlapping feature matching pairs that meet the area standards; and taking the first-ranked target overlapping feature matching pair as the target reference point.

[0016] Preferably, the feature simplification of the initial three-dimensional model and the monitoring three-dimensional model to obtain the first simplified model and the second simplified model respectively includes: based on the attention mechanism, the K-Means clustering algorithm and the target reference point, the slope feature recognition is performed on the initial setting information and the monitoring images of each angle, and the redundant features in the initial three-dimensional model and the monitoring three-dimensional model are eliminated to obtain the first simplified model and the second simplified model in a three-dimensional state with slope feature information.

[0017] Preferably, before obtaining the monitoring images of various angles taken by the drone at low altitude and the corresponding initial setting information, a high-precision measurement method for three-dimensional displacement of highway facilities based on low altitude also includes: sending slope displacement measurement information to the corresponding base station control center according to preset slope image data acquisition rules; the base station control center determines the low-altitude path planning strategy and image acquisition strategy of the drone based on the acquired drone status information and slope displacement measurement information.

[0018] Preferably, the sending of slope displacement measurement information to the corresponding base station control center according to preset slope image data acquisition rules includes: acquiring environmental status information and traffic information of the measurement area; extracting features of the environmental status information and traffic information according to preset standards to obtain acquisition frequency influencing features; analyzing each acquisition frequency influencing feature and slope status information based on a fourth preset analysis rule to determine the acquisition frequency, acquisition time and acquisition position of the monitoring image; and using the acquisition frequency, acquisition time and acquisition position as the slope displacement measurement information.

[0019] Preferably, the fourth preset analysis rule is used to analyze the influencing features of each acquisition frequency and the slope status information to determine the acquisition frequency, acquisition time and acquisition location of the monitoring image, including: using a multi-head attention mechanism algorithm to analyze the influencing features of each acquisition frequency to determine the first impact rate value of the slope change of each acquisition frequency impact feature; using a feature fusion algorithm to fuse the features of each acquisition frequency impact, and estimating the second impact rate value of the slope based on the fused features and the slope status information; determining the acquisition frequency, acquisition time and acquisition location of the monitoring image based on the slope status information, the first impact rate value and the second impact rate value.

[0020] The beneficial effects of the present invention are:

[0021] 1. The present invention uses an unmanned aerial vehicle carrying a high-definition camera to shoot monitoring images of the slope at a low-altitude fixed position, performs displacement analysis on the reference target in the monitoring image, and determines the displacement change degree of the reference target. When the first displacement change value of the reference target is less than the first preset value, the monitoring image is analyzed using the first preset analysis rule to determine the displacement value of each target on each axis in the three-dimensional coordinate system; when the first displacement change value is greater than or equal to the first preset value, the target reference point is first established, and then a three-dimensional coordinate system is formed based on the target reference point, and then the displacement change value of each target is analyzed to determine the displacement value of each target. In the above manner, the present invention solves the problem that the traditional method can only obtain data in two-dimensional directions, and multiple inclinometers need to be set up, the construction is more complicated, and the later maintenance cost is high.

[0022] 2. The present invention first analyzes the displacement of the reference target, and selects different slope displacement analysis methods according to different analysis results of the reference target to determine the displacement of the slope in three directions, which can avoid the problem of low accuracy of the slope displacement measurement results when the reference target changes; at the same time, according to the actual analysis results, different analysis methods are selected to improve the accuracy of the slope displacement measurement; and the displacement change of the slope is analyzed from the perspective of the three-dimensional coordinate system, which can improve the accuracy of the slope displacement measurement.

[0023] 3. When the reference target loses its reference function, the present invention adopts a method of constructing a monitoring three-dimensional model of the current slope state based on the monitoring image; performing comparative analysis on the initial three-dimensional model and the monitoring three-dimensional model to determine the target reference point; performing feature simplification on the initial three-dimensional model and the monitoring three-dimensional model to obtain a first simplified model and a second simplified model respectively; overlapping the target reference points in the first simplified model and the second simplified model, and then determining the displacement value of the slope according to the first preset analysis rule. Through the above method, the present invention can determine the change value of the slope displacement by constructing a three-dimensional model for precise comparative analysis, and when the reference target loses its function, it can still determine the change in the slope displacement, which reflects the robustness of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 The present invention is a schematic flow chart of a method for high-precision measurement of three-dimensional displacement of highway facilities based on low altitude. DETAILED DESCRIPTION

[0025] The following will refer to the attached Figure 1 The embodiments of the present invention are described in detail. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0026] In one embodiment of the present invention, a high-precision method for measuring the three-dimensional displacement of highway facilities based on low altitude is provided. Figure 1 As shown, the following steps are included:

[0027] Step S11: Acquire monitoring images of various angles taken by the drone at low altitude and the corresponding initial setting information.

[0028] Specifically, when taking low-altitude images of the slope using a drone, the drone carries a precise positioning device, and when it moves to the set coordinates, it will take pictures to obtain monitoring images; the initial setting information includes the initial setting images of various angles, the initial three-dimensional model obtained based on the initial setting images, the measurement mapping amounts of each measurement target on each axis of the three-dimensional coordinate system, etc.; the initial setting image is the monitoring image obtained by taking the drone for the first time after the targets on the slope are set.

[0029] Furthermore, after shooting, the monitoring image will be transmitted to the back-end monitoring personnel to determine whether there is obvious abnormal information in the monitoring image. If not, the control information of the drone will not be changed; if there is obvious abnormal information, such as pedestrians or animals on the slope, a second monitoring image will be collected at intervals to improve the accuracy of slope displacement measurement.

[0030] Step S12: Analyze the state of the reference target according to the monitoring image to determine a first displacement change value of the reference target.

[0031] This step is used to determine the displacement change of the reference target. When the displacement change does not meet the preset standard, the reference target is no longer suitable as a reference point, and subsequent displacement analysis of the target is performed, that is, a new analysis method needs to be reselected, and step S14 is used to analyze the slope displacement; when the displacement change meets the preset standard, step S13 is used to perform the displacement analysis of the slope.

[0032] Step S13: When the first displacement change value is less than the first preset value, the monitoring image is analyzed and processed based on the first preset analysis rule to determine the displacement value of each target.

[0033] In step S13, the first preset analysis rule includes constructing a three-dimensional model based on the monitoring image, forming a three-dimensional coordinate system based on the reference target features in the three-dimensional model; then determining the measurement mapping amount of each measurement target on the x, y and z axes based on the three-dimensional coordinate system; and analyzing the measurement mapping amount based on the initial mapping amount to determine the displacement value of each target on each axis.

[0034] Step S14: When the first displacement change value is greater than or equal to the first preset value, an initial three-dimensional model constructed by initial setting information is obtained, and a monitoring three-dimensional model of the current slope state is constructed according to the monitoring image; the initial three-dimensional model and the monitoring three-dimensional model are compared and analyzed to determine the target reference point; the initial three-dimensional model and the monitoring three-dimensional model are feature simplified to obtain a first simplified model and a second simplified model respectively; the target reference points in the first simplified model and the second simplified model are overlapped to determine the displacement value of the slope according to the first preset analysis rule.

[0035] In this embodiment, the present invention uses an unmanned aerial vehicle carrying a high-definition camera to shoot monitoring images of the slope at a low-altitude fixed position, performs displacement analysis on the reference target in the monitoring image, and determines the displacement change degree of the reference target. When the first displacement change value of the reference target is less than the first preset value, the monitoring image is analyzed using the first preset analysis rule to determine the displacement value of each target on each axis in the three-dimensional coordinate system; when the first displacement change value is greater than or equal to the first preset value, the target reference point is first established, and then a three-dimensional coordinate system is formed based on the target reference point, and then the displacement change value of each target is analyzed to determine the displacement value of each target in the direction of each axis of the three-dimensional coordinate system. In the above manner, the present invention solves the problem that the traditional method can only obtain data in the two-dimensional direction, and multiple inclinometers need to be set up, the construction is more complicated, and the later maintenance cost is high.

[0036] Furthermore, the present invention first analyzes the displacement of the reference target, and selects different slope displacement analysis methods according to different analysis results of the reference target to determine the displacement of the slope in three directions, which can avoid the problem of low accuracy of the slope displacement measurement results when the reference target changes; at the same time, according to the actual analysis results, different analysis methods are selected to improve the accuracy of the slope displacement measurement; and the displacement change of the slope is analyzed from the perspective of the three-dimensional coordinate system, which can improve the accuracy of the slope displacement measurement.

[0037] Furthermore, when the reference target loses its reference function, the present invention adopts a method of constructing a monitoring three-dimensional model of the current slope state based on the monitoring image; performing comparative analysis on the initial three-dimensional model and the monitoring three-dimensional model to determine the target reference point; performing feature simplification on the initial three-dimensional model and the monitoring three-dimensional model to obtain a first simplified model and a second simplified model respectively; overlapping the target reference points in the first simplified model and the second simplified model, and then determining the displacement value of the slope according to the first preset analysis rule. Through the above method, the present invention can determine the change value of the slope displacement by constructing a three-dimensional model for accurate comparative analysis, and when the reference target loses its function, it can still determine the change in the slope displacement, which reflects the robustness of the present invention in slope displacement measurement.

[0038] In one embodiment of the present invention, a high-precision measurement method for three-dimensional displacement of highway facilities based on low altitude also includes: when a first displacement change value is greater than or equal to a first preset value, performing a comparison analysis based on initial setting information and a monitoring image to determine a second reference target; according to the second reference target, analyzing and processing the monitoring image based on a first preset analysis rule to determine the displacement value of each target.

[0039] In this embodiment, a two-dimensional feature point analysis is performed based on the initial setting images at each angle in the initial setting information and the monitoring images taken at the same position to determine the feature points that have not changed in each monitoring image; and then a second reference target that can form a three-dimensional coordinate system is selected from the second feature points. After determining the second reference target, the monitoring image is processed using the first preset analysis rule to determine the displacement value of each target mapped on each axis of the three-dimensional coordinate system. Through the setting method of this embodiment, the present invention can compare and analyze the initial setting information and the monitoring image to determine the second reference target that can form a three-dimensional coordinate system when the reference coordinates lose their measurement accuracy; the displacement value of each target is determined based on the second reference target, which helps to improve the accuracy and measurement precision of the measurement results.

[0040] In one embodiment of the present invention, the state of the reference target is analyzed according to the monitoring image to determine the first displacement change value of the reference target, including: preprocessing each monitoring image, identifying the reference target and extracting features to obtain the reference target features; matching and analyzing the initial target features with the reference target features, overlapping one of the reference target features in the same direction, and calculating the offset value of the initial target feature and the reference target feature; and using the two offset values ​​as the first displacement change value.

[0041] In this embodiment, the preprocessing process includes grayscale processing, noise reduction processing and edge enhancement, so as to improve the accuracy of target identification during target identification; after identifying the reference target, extract the reference target features; then match and analyze the initial target features with the reference target features, that is, overlap one of the reference target features in the same direction, calculate the offset values ​​of the other two initial target features and the reference target features; use these two offset values ​​as the first displacement change value. The first displacement change value includes two values, and there is no limitation on which axis of the three-dimensional coordinate system the offset value belongs to. The first preset value is the corresponding evaluation value of the offset value, and the evaluation values ​​of the two offset values ​​are the same.

[0042] Through the setting method of this embodiment, the present invention can analyze the state of the reference target according to the monitoring image, accurately determine the first displacement change value of the reference target, and thus select different slope displacement analysis methods according to the relationship between the first change value and the first preset value, which helps to improve the robustness and measurement accuracy of the slope displacement measurement.

[0043] In one embodiment of the present invention, the monitoring images are analyzed and processed based on a first preset analysis rule to determine the displacement value of each target, including: performing target recognition and feature extraction on each monitoring image, and constructing a three-dimensional model based on the extracted features; forming a three-dimensional coordinate system with reference target features; determining the measurement mapping amount of each measurement target on each axis based on the three-dimensional coordinate system; and analyzing the measurement mapping amount based on the initial mapping amount to determine the displacement value of each target.

[0044] Through the configuration of this embodiment, the present invention can determine the displacement value of the slope from three directions of a three-dimensional coordinate system. Compared with the traditional method that can only obtain data in two-dimensional directions, the present invention can improve the measurement accuracy of the slope displacement.

[0045] In one embodiment of the present invention, a comparative analysis is performed on the initial three-dimensional model and the monitoring three-dimensional model to determine the target reference point, including: overlapping the initial three-dimensional model and the monitoring three-dimensional model according to the initially set direction to obtain a three-dimensional model formed by overlapping feature points; performing a reference point analysis of the overlapping feature points in a three-dimensional coordinate system to obtain overlapping feature matching pairs; identifying and analyzing the object areas corresponding to the overlapping feature matching pairs to determine target overlapping feature matching pairs that meet the area standards; and taking the first-ranked target overlapping feature matching pair as the target reference point.

[0046] Through the configuration of this embodiment, the present invention can select a suitable target coincidence feature matching pair according to the object area corresponding to the coincidence feature, and use the target coincidence feature matching pair with the smallest area as the target reference point, so as to improve the measurement accuracy of the slope displacement.

[0047] In one embodiment of the present invention, the initial three-dimensional model and the monitoring three-dimensional model are feature simplified to obtain a first simplified model and a second simplified model respectively, including: performing slope feature recognition on the initial setting information and the monitoring images at various angles based on the attention mechanism, the K-Means clustering algorithm and the target reference point, and eliminating redundant features in the initial three-dimensional model and the monitoring three-dimensional model to obtain the first simplified model and the second simplified model in a three-dimensional state with slope feature information.

[0048] The specific process is: use the K-Means clustering algorithm combined with the attention mechanism to perform cluster analysis on the pixels corresponding to the initial setting image in the initial setting information, so as to separate the images corresponding to the slope, benchmark target, and target benchmark point; then form a key three-dimensional model according to the stripped image according to the original position, and process the initial three-dimensional model according to the feature range of the key three-dimensional model, and eliminate other data outside the range to obtain a first streamlined model of the three-dimensional form.

[0049] Similarly, the process of performing feature simplification on the monitoring three-dimensional model to obtain the second simplified model is the same as the above method.

[0050] Through the setting method of this embodiment, the present invention can simplify the features of the initial three-dimensional model and the monitoring three-dimensional model, remove redundant features in the model, and obtain a first simplified model and a second simplified model respectively, so that when the target reference points in the first simplified model and the second simplified model are subsequently overlapped, the displacement value of the slope can be determined according to the first preset analysis rule. The displacement value of the slope can be determined with high precision and the speed of data processing can be improved.

[0051] In one embodiment of the present invention, before obtaining the monitoring images of various angles taken by the drone at low altitude and the corresponding initial setting information, it also includes: sending slope displacement measurement information to the corresponding base station control center according to the preset slope image data acquisition rules; the base station control center determines the low-altitude path planning strategy and image acquisition strategy of the drone based on the acquired drone status information and slope displacement measurement information.

[0052] Furthermore, in one embodiment of the present invention, according to preset slope image data acquisition rules, slope displacement measurement information is sent to the corresponding base station control center, including: obtaining environmental status information and traffic information of the measurement area; extracting features of the environmental status information and traffic information according to preset standards to obtain acquisition frequency influencing features; analyzing each acquisition frequency influencing feature based on a fourth preset analysis rule to determine the acquisition frequency, acquisition time and acquisition position of the monitoring image; and using the acquisition frequency, acquisition time and acquisition position as the slope displacement measurement information.

[0053] Preferably, the environmental status information includes weather temperature, humidity, distance to water source, soil status around the slope, depth of ditch beside the slope, etc., and the traffic information includes different types of vehicle flow, etc.

[0054] The preset standard is a set threshold or set interval corresponding to each environmental state information and traffic information; when the collected environmental state information and traffic information meet the preset standard, the type and value of the environmental state information or traffic information are extracted as the collection frequency influencing features, and then the collection frequency of the slope monitoring image is analyzed according to these influencing features. Through the setting method of this embodiment, the present invention can flexibly determine the collection frequency of obtaining the slope monitoring image according to the environmental state information and traffic information around the slope, and then can timely determine the displacement state of the slope. When the slope displacement exceeds the set value, it is added to the list of slopes to be repaired, and the monitoring personnel are reminded so that the slope can be repaired in time to ensure the safety of road driving.

[0055] In one embodiment of the present invention, each acquisition frequency influencing feature and slope status information are analyzed based on a fourth preset analysis rule to determine the acquisition frequency, acquisition time and acquisition location of the monitoring image, including: using a multi-head attention mechanism algorithm to analyze each acquisition frequency influencing feature to determine the first impact rate value of the slope change of each acquisition frequency influencing feature; using a feature fusion algorithm to perform feature fusion on each acquisition frequency influencing feature, and estimating the second impact rate value of the slope based on the fused features and the slope status information; determining the acquisition frequency, acquisition time and acquisition location of the monitoring image based on the slope status information, the first impact rate value and the second impact rate value.

[0056] In this embodiment, the present invention adopts a multi-head attention mechanism algorithm to analyze the influencing features of each acquisition frequency, determine the key abnormal features for attention; and assign a higher weight to the influencing features of attention, so that the first influencing rate value of the slope change is more realistic and the first influencing rate value obtained is more accurate. And the feature fusion algorithm is used to fuse the associated features in the influencing features of each acquisition frequency to obtain the fusion feature; and then the second influencing rate value of the slope displacement change is determined according to the fusion feature and the current slope state information. Finally, the weighted average method is used to analyze the first influencing rate value and the second influencing rate value to determine the average influencing rate, and the acquisition frequency is determined according to the corresponding relationship between the average influencing rate and the acquisition frequency; the acquisition time and acquisition position are determined according to the current slope state information. Through the setting method of this embodiment, the present invention can dynamically and reasonably adjust the acquisition frequency of the drone to obtain the slope image and monitor the state of the slope; when the slope displacement exceeds the preset value, it is repaired in time to ensure the safety of road driving.

[0057] In one embodiment of the present invention, the base station control center determines the low-altitude path planning strategy and image acquisition strategy of the drone based on the acquired drone status information and slope displacement measurement information, including: after receiving the image acquisition information of the slope, the base station control center of the drone base station will initialize the status information of each drone, and when the status information meets the conditions for drone flight and image acquisition, the drone is added to the acquisition queue; after all drone status information is obtained, according to the slope position and drone acquisition position to be measured, the flight path control information, image acquisition position information and image acquisition angle information are sent to the drones that meet the conditions; then the drone flies according to the flight path information, and automatically avoids obstacles when obstacles are detected, until the acquisition task is completed and returns to the drone base station.

[0058] Furthermore, when the UAV completes the image acquisition task and there are no qualified UAVs in the base station, the base station control center analyzes the flight path according to the current position of the UAV and the target slope position; then the base station control center sends the image acquisition angle and position to the UAV, controls the UAV to fly to the specified position, and collects images of the slope at various angles.

[0059] In one embodiment of the present invention, the central control system sends a slope displacement measurement instruction to the base station control center at the corresponding position according to the preset slope image data acquisition rules; when the base station control center receives the slope displacement measurement instruction, it obtains the status information of each drone in the drone base station, analyzes each status information, acquisition instruction and acquisition position to determine the low-altitude path planning strategy and image acquisition strategy based on the drone; during the flight of the drone, automatic obstacle avoidance is completed until the designated slope image acquisition position is reached and image data of each angle is acquired according to the setting requirements; monitoring images of each angle taken by the drone at low altitude and the corresponding initial setting information are acquired; the displacement change degree of the reference target is judged, and when the displacement change degree of the reference target is less than the first preset value, the monitoring image is processed to construct a three-dimensional model of each target; displacement analysis of each target is performed according to the three-dimensional model to determine the displacement value in each direction of the slope.

[0060] When the displacement change degree of the reference target is greater than or equal to the first preset value, a comparison analysis is performed based on the initial setting information and the monitoring image to determine the second reference target; according to the second reference target, the monitoring image is analyzed to construct a three-dimensional point cloud model, and the displacement value of each target is determined based on the three-dimensional point cloud model.

[0061] Alternatively, when the displacement change degree of the reference target is greater than a preset value, an initial three-dimensional model constructed by initial setting information is obtained, and a monitoring three-dimensional model of the current slope state is constructed according to the monitoring image; the initial three-dimensional model and the monitoring three-dimensional model are compared and analyzed to determine the target reference point; the initial three-dimensional model and the monitoring three-dimensional model are feature simplified to obtain a first simplified model and a second simplified model respectively; the target reference points in the first simplified model and the second simplified model are overlapped to determine the displacement value of the slope according to a first preset analysis rule.

[0062] In this embodiment, the present invention uses a drone carrying a high-definition camera to shoot monitoring images of the slope at a low altitude fixed position, performs displacement analysis on the reference target in the monitoring image, and determines the displacement change degree of the reference target. When the first displacement change value of the reference target is less than the first preset value, the monitoring image is analyzed using the first preset analysis rule to determine the displacement value of each target on each axis in the three-dimensional coordinate system; when the first displacement change value is greater than or equal to the first preset value, the target reference point is first established, and then a three-dimensional coordinate system is formed based on the target reference point, and then the displacement change value of each target is analyzed to determine the displacement value of each target. In the above manner, the present invention solves the problem that the traditional method can only obtain data in two-dimensional directions, and multiple inclinometers need to be set up, the construction is complex, and the later maintenance cost is high.

[0063] In one embodiment of the present invention, a high-precision measurement system for three-dimensional displacement of highway facilities based on low altitude includes an unmanned aerial vehicle and a base station control center, a set target, and a solar power supply system that provides energy for the unmanned aerial vehicle and the base station. The high-precision measurement system for three-dimensional displacement of highway facilities based on low altitude of the present invention can conveniently, intuitively, and efficiently monitor the safety of slopes. The unmanned aerial vehicle base station is used to provide a reliable communication connection and ensure the safe and efficient flight of the unmanned aerial vehicle. The solar power supply system is used as a power source, which is convenient and environmentally friendly while also ensuring the endurance of the unmanned aerial vehicle. The target, including the reference target and the measurement target, is used to provide a reference for whether the slope has been displaced.

[0064] Among them, three reference targets are installed at stable positions near the slope as reference points to determine the three-dimensional coordinate system of the slope. According to the specific conditions of the slope, a corresponding number of measurement targets are set on the slope surface as the main shooting objects of the aerial photography.

[0065] Assume that the initial coordinates of the drone's aerial photography point are (x0, y0, z0), the angles between the aerial photography point and the target point and the x-axis, y-axis, and z-axis of the coordinate system are θ1, θ2, and θ3 respectively, and the length of the line is L; the real-time coordinates of the target point are (x0+Lcosθ1, y0+Lcosθ2, z0+Lcosθ3); the mapping changes of the line in the x-axis, y-axis, and z-axis directions are Lcosθ1, Lcosθ2, and Lcosθ3 respectively.

[0066] Among them, the aerial photography point is the aerial photography position of the drone camera, and the target point is the center position of the measurement target.

[0067] Through the above method, the present invention can intuitively reflect the displacement of the slope in all directions, and the obtained measurement data is more comprehensive and accurate. In addition, the present invention can be applied to a variety of highway slope environments, is easy to construct and operate, can greatly reduce labor costs, and has high promotion value.

[0068] In one embodiment of the present invention, the content of constructing a monitoring 3D model of the current slope state based on the monitoring image includes: using a multi-view reconstruction method to process 2D images taken from multiple different perspectives to form a slope state map in a 3D scene, and the slope state map includes 3D images of the slope, reference targets and other objects. In order to achieve accurate and efficient multi-view reconstruction, a series of optimization algorithms are used to improve reconstruction quality and computational efficiency.

[0069] More robust feature point extraction operators are used: such as SIFT (Scale Invariant Feature Transform) and SURF (Speeded Robust Features). These operators have good robustness to factors such as image rotation, scale change, and illumination change. They can extract stable and accurate feature points under complex image conditions and improve the accuracy of feature point matching in multi-view reconstruction. In addition, a fast nearest neighbor search library combined with a K-nearest neighbor algorithm is used to quickly find the most similar feature point pairs in a large-scale feature point set, thereby speeding up feature point matching. In addition, multiple constraints such as the descriptor similarity, geometric position relationship, and camera pose estimation of feature points are comprehensively considered to reduce the occurrence of mismatches and improve the accuracy of matching. In addition, feature point pairs are randomly extracted to estimate the camera model parameters carried by the drone, and mismatched points are removed according to a certain number of iterations and inlier thresholds, thereby improving the accuracy and stability of camera pose estimation.

[0070] The rotation and translation parameters of the camera are solved by minimizing the projection error of the feature points on the image plane, so that the information of the feature points can be fully utilized to obtain more accurate pose estimation results.

[0071] Utilize the geometric constraints of the scene: If some geometric information in the scene is known, such as coplanarity, colinearity, symmetry and other relationships, these constraints can be added to the pose estimation process to reduce the number of unknown parameters and improve the accuracy and reliability of pose estimation.

[0072] After obtaining the initial 3D reconstruction model, mesh simplification algorithms, such as the QEM (Quadratic Error Metrics) algorithm and Vertex Decimation, are used to simplify the model, reduce the number of faces and vertices of the model, and try to maintain the shape and detail features of the model to improve the processing efficiency and visualization effect of the model. The surface of the reconstructed 3D model is smoothed to remove noise and tiny irregularities, making the model surface smoother and more natural. Filtering algorithms, such as Gaussian filtering and bilateral filtering, are used to smooth the vertex coordinates of the model.

[0073] According to the characteristics and requirements of multi-view reconstruction, select appropriate data structures to store and process information such as images, feature points, and camera parameters. For example, use data structures such as octrees and KD trees to speed up the search and matching operations of feature points. And analyze and optimize the key algorithms in multi-view reconstruction to reduce unnecessary calculation steps and data access times. For example, in the camera pose estimation algorithm, use more efficient linear algebra calculation methods and optimization strategies to reduce the time complexity of the algorithm.

[0074] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0075] It should be noted that, in the description of the present invention, the terms "first", "second" and "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0076] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, implements the functions / operations specified in the flow chart and / or block diagram. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0077] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0078] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0079] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0080] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0081] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A high-precision measurement method for three-dimensional displacement of highway facilities based on low altitude, characterized in that: include: Obtain monitoring images of various angles taken by drones at low altitudes and the corresponding initial setting information; Analyzing the state of the reference target according to the monitoring image to determine a first displacement change value of the reference target; When the first displacement change value is less than the first preset value, analyzing and processing the monitoring image based on the first preset analysis rule to determine the displacement value of each target; When the first displacement change value is greater than or equal to the first preset value, an initial three-dimensional model constructed by initial setting information is obtained, and a monitoring three-dimensional model of the current slope state is constructed according to the monitoring image; the initial three-dimensional model and the monitoring three-dimensional model are compared and analyzed to determine the target reference point; The initial three-dimensional model and the monitoring three-dimensional model are feature simplified to obtain a first simplified model and a second simplified model respectively; the target reference points in the first simplified model and the second simplified model are overlapped to determine the displacement value of the slope according to a first preset analysis rule.

2. The high-precision measurement method for three-dimensional displacement of highway facilities according to claim 1 is characterized in that: Also includes: When the first displacement change value is greater than or equal to the first preset value, performing a comparison analysis based on the initial setting information and the monitoring image to determine a second reference target; According to the second reference target, the monitoring image is analyzed and processed based on the first preset analysis rule to determine the displacement value of each target.

3. The high-precision measurement method for three-dimensional displacement of highway facilities according to claim 1 is characterized in that: The method of analyzing the state of the reference target according to the monitoring image to determine the first displacement change value of the reference target includes: preprocessing each monitoring image, identifying the reference target and extracting features to obtain the reference target features; performing matching analysis based on the initial target features and the reference target features, overlapping one of the reference target features in the same direction, and calculating the offset value of the initial target feature and the reference target feature; and using the two offset values ​​as the first displacement change value.

4. The high-precision measurement method for three-dimensional displacement of highway facilities according to claim 1 is characterized in that: The method of analyzing and processing the monitoring images based on the first preset analysis rule to determine the displacement value of each target includes: performing target recognition and feature extraction on each monitoring image, and constructing a three-dimensional model based on the extracted features; forming a three-dimensional coordinate system based on the reference target features; determining the measurement mapping amount of each measurement target on each axis based on the three-dimensional coordinate system; and analyzing the measurement mapping amount based on the initial mapping amount to determine the displacement value of each target.

5. The high-precision measurement method for three-dimensional displacement of highway facilities according to claim 1 is characterized in that: The comparative analysis of the initial three-dimensional model and the monitoring three-dimensional model to determine the target reference point includes: overlapping the initial three-dimensional model and the monitoring three-dimensional model according to the initial set direction to obtain a three-dimensional model formed by overlapping feature points; performing a reference point analysis of the overlapping feature points in a three-dimensional coordinate system to obtain overlapping feature matching pairs; identifying and analyzing the object areas corresponding to the overlapping feature matching pairs to determine the target overlapping feature matching pairs that meet the area standards; and taking the first-ranked target overlapping feature matching pair as the target reference point.

6. The high-precision measurement method for three-dimensional displacement of highway facilities according to claim 1 is characterized in that: The feature simplification of the initial three-dimensional model and the monitoring three-dimensional model to obtain the first simplified model and the second simplified model respectively includes: based on the attention mechanism, the K-Means clustering algorithm and the target reference point, the initial setting information and the monitoring images of each angle are subjected to slope feature recognition, and the redundant features in the initial three-dimensional model and the monitoring three-dimensional model are eliminated to obtain the first simplified model and the second simplified model in a three-dimensional state with slope feature information.

7. The high-precision measurement method for three-dimensional displacement of highway facilities according to claim 1 is characterized in that: Before obtaining the monitoring images of various angles taken by the drone at low altitude and the corresponding initial setting information, it also includes: sending slope displacement measurement information to the corresponding base station control center according to the preset slope image data acquisition rules; the base station control center determines the low-altitude path planning strategy and image acquisition strategy of the drone based on the acquired drone status information and slope displacement measurement information.

8. The high-precision measurement method for three-dimensional displacement of highway facilities according to claim 7 is characterized in that: The method of sending slope displacement measurement information to a corresponding base station control center according to a preset slope image data acquisition rule includes: obtaining environmental status information and traffic information of a measurement area; extracting features of the environmental status information and traffic information according to preset standards to obtain acquisition frequency influencing features; analyzing each acquisition frequency influencing feature and slope status information based on a fourth preset analysis rule to determine the acquisition frequency, acquisition time and acquisition position of the monitoring image; and using the acquisition frequency, acquisition time and acquisition position as the slope displacement measurement information.

9. The high-precision measurement method for three-dimensional displacement of highway facilities according to claim 8 is characterized in that: The method of analyzing the influencing features of each acquisition frequency and the slope status information based on the fourth preset analysis rule to determine the acquisition frequency, acquisition time and acquisition position of the monitoring image includes: using a multi-head attention mechanism algorithm to analyze the influencing features of each acquisition frequency to determine the first impact rate value of the slope change of each acquisition frequency impact feature; using a feature fusion algorithm to fuse the features of each acquisition frequency impact, and estimating the second impact rate value of the slope based on the fused features and the slope status information; determining the acquisition frequency, acquisition time and acquisition position of the monitoring image based on the slope status information, the first impact rate value and the second impact rate value.

Images