Linear engineering slope disease detection method, system and equipment
Through the drone collecting multi-contour information and building a three-dimensional digital geological model, combining convolutional neural network to identify slope diseases and calculate safety coefficients, the problems of low patrol efficiency and high safety risks in the existing technology are solved, and linearity is achieved Efficiency and safety assessment of engineering slopes.
Patent Information
- Application Number
- CN202411050057.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-01
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-08-01
AI Technical Summary
In the prior art, the inspection efficiency of linear engineering slopes is low and there is a safety risk, especially the inability to effectively inspect steep slopes, resulting in blind spots.
UAVs are used to collect multi-contour information, and a three-dimensional digital geological model of slopes is constructed through aerial images, depth images and laser detection point clouds. It combines with convolutional neural networks to identify diseases, and calculates slope safety coefficients to evaluate stability.
It significantly improves the efficiency and accuracy of slope disease inspection, reduces safety risks, and achieves effective assessment of steep slopes.
Smart Images

Figure CN118941539A_ABST
Abstract
Description
Background Art
[0002] Due to the linear characteristics of railway / highway transportation infrastructure, a large number of transportation engineering construction will inevitably produce a huge number of deep-cut road slopes. According to incomplete statistics, the stock of various types of railway / highway slopes in service in my country has exceeded hundreds of thousands, and a large number of slopes are "high, steep and dangerous". With the further development and improvement of my country's transportation infrastructure construction, there will be more linear engineering slopes in operation in the future, and their safe operation will become a key node to ensure traffic safety.
[0003] To ensure the safety of the slopes of linear projects during the operation period and actively carry out the work of disaster prevention and mitigation of geological disasters on the slopes, we should start from two aspects: first, increase the intensity of inspections to effectively identify the form and scale of slope disasters; second, do a good job in the stability evaluation of diseased slopes and understand their safety status. At present, the safety inspection of linear project slopes in my country mainly adopts the method of manual inspection, and the inspection results are manually recorded or photographed by visual, knocking, touching and other methods. There are two significant disadvantages of traditional manual inspections: (1) In areas with large undulating slope terrain, manual inspections are time-consuming and labor-intensive, and extremely dangerous, resulting in safety hazards in the inspection process; (2) For steep slopes, inspection personnel cannot reach them, and walking inspections cannot meet work requirements, resulting in blind spots in the inspection process. Therefore, in response to this situation, the present invention constructs an integrated technology of "intelligent identification → three-dimensional reconstruction → rapid evaluation" of linear project slope diseases, which can significantly improve the inspection efficiency of slope diseases and has important scientific significance and engineering value for ensuring the safe operation and maintenance of linear projects. Summary of the invention
[0004] In order to solve the above problems in the prior art, namely, the technical problems of low efficiency and high safety risk of existing manual slope inspection and assessment, the present invention provides a linear engineering slope disease detection method, system and equipment.
[0005] The present invention provides a linear engineering slope disease detection method, which adopts the following technical scheme: Collecting multivariate contour information of the target area; the multivariate contour information includes aerial images, depth images and laser detection point clouds; Aligning the multivariate contour information, and setting the multivariate contour information category weights according to the aerial image; Construct a three-dimensional digital geological model of the slope based on weighted multivariate contours; Extracting geometric parameters and geological parameters of a designated two-dimensional slope surface according to the three-dimensional digital geological model of the slope; Based on the geometric and geological parameters, calculate the slope safety factor As a slope safety factor of a specified two-dimensional slope surface, the stability of the slope is evaluated by the slope safety factor; The slope safety factor for: ; represents the safety factor; It indicates the soil bulk density; Indicates adhesion; represents the friction angle; Indicates slope height; represents the downhill angle; From the slope uphill angle Determined undetermined constants; The slope downslope angle Determine the undetermined constant.
[0006] Preferably, the method for acquiring the multivariate profile information is: The information collection equipment continuously collects the slope profile information of each dimension in the target area according to the respective set routes; The slope profile information of different dimensions is aligned and / or data fused through preset markers or automatically identified key points to obtain multivariate profile information.
[0007] Preferably, the specified two-dimensional slope is selected by: Extracting slope image features based on the aerial image; Based on the slope image features, a slope disease recognition result is obtained through a disease recognition network based on a convolutional neural network; The two-dimensional slope surface with slope disease identification results indicating that it has disease is selected as the designated two-dimensional slope surface.
[0008] Preferably, the training method of the disease recognition network based on convolutional neural network includes: Perform feature extraction and cluster analysis on the labeled aerial images to obtain the feature parameters of the historical diseased slope images and the feature parameters of the historical disease-free slope images; The characteristic parameters of the diseased slope image and the characteristic parameters of the historical disease-free slope image are mixed and passed through the disease recognition network based on the convolutional neural network to be trained to obtain the slope disease recognition result of the training set, the model parameters are adjusted and iterated to the preset termination condition, and the trained disease recognition network based on the convolutional neural network is obtained.
[0009] Preferably, the construction of the three-dimensional digital geological model of the slope includes: Based on the characteristic parameters of multivariate contour information, the three-dimensional reconstruction model of the slope is obtained through the preset image matching and motion recovery structure algorithm; The three-dimensional reconstructed model of the slope is converted into a point cloud data model, and the point cloud data model is densely reconstructed based on a depth map fusion method to obtain an elevation cloud map model of the slope; According to the position information of the dense points in the elevation cloud model, the elevation cloud model is discretized to obtain a three-dimensional digital model of the diseased slope at the site scale.
[0010] Preferably, feature extraction and cluster analysis are performed on the aerial images with labels: Based on the historical aerial images in the initial database of slope diseases, different characteristic parameter combinations are determined; Clustering the historical aerial images according to the different characteristic parameter combinations to obtain clustering results; According to the clustering result, characteristic parameters of historical diseased slope images are screened out from historical aerial images.
[0011] Another aspect of the present invention provides a linear engineering slope disease detection system, the system comprising: A multivariate contour information acquisition module is used to collect multivariate contour information of a target area through different sensors; the multivariate contour information includes an aerial image, a depth image, and a laser detection point cloud; A data alignment module, used for aligning the multivariate profile information and setting the multivariate profile information category weights according to the aerial image; The diseased slope 3D model reconstruction module is used to construct a 3D digital geological model of the slope based on weighted multivariate contours; A parameter extraction module, used for extracting geometric parameters and geological parameters of a designated two-dimensional slope surface according to the three-dimensional digital geological model of the slope; Diseased slope stability assessment module, used for Based on the geometric and geological parameters, calculate the slope safety factor As a slope safety factor of a specified two-dimensional slope surface, the stability of the slope is evaluated by the slope safety factor; The slope safety factor for: ; represents the safety factor; It indicates the soil bulk density; Indicates adhesion; represents the friction angle; Indicates slope height; represents the downhill angle; From the slope uphill angle Determined undetermined constants; The slope downslope angle Determine the undetermined constant.
[0012] Preferably, the diseased slope three-dimensional model reconstruction module, the diseased slope stability rapid assessment module and the image information transmission and acquisition module are integrated on an indoor server platform.
[0013] According to a third aspect of the present invention, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to at least one of the processors; wherein, The memory stores instructions that can be executed by the processor, and the instructions are used to be executed by the processor to implement the above-mentioned linear engineering slope disease detection method.
[0014] In a fourth aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to be executed by the computer to implement the above-mentioned linear engineering slope disease detection method.
[0015] Beneficial effects of the present invention: (1) The present invention uses drone aerial photography to conduct safety inspections on linear engineering slopes, effectively reducing the safety risks of the original manual inspections on steep slopes and improving the efficiency of slope safety inspections.
[0016] (2) The present invention constructs an integrated technology of "intelligent identification → three-dimensional reconstruction → rapid evaluation" of linear engineering slope diseases. By training the disease recognition network, a three-dimensional digital model of the diseased slope is constructed, and the slope safety factor is calculated based on the obtained geometric parameters and geological parameters. Compared with the previous manual recording or photography of inspection results by visual inspection, knocking, touch, etc., the accuracy and efficiency of slope disease assessment are greatly improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 It is a flow chart of the linear engineering slope disease detection system evaluation method of the present application; Figure 2 This is a schematic diagram of a linear engineering slope disease detection system according to an embodiment of the present application; Figure 3 is a schematic diagram of a two-dimensional section of a slope involved in this application; Figure 4 This is a flow chart of the intelligent evaluation method for linear engineering slope diseases with specified two-dimensional slope confirmation added to the application; Description of reference numerals: 1. Slope disease intelligent identification module; 2. Diseased slope three-dimensional model reconstruction module; 3. Diseased slope stability rapid assessment module; 4. Image information transmission and acquisition module; 5. Indoor server platform; 6. Unmanned aerial vehicle; 7. Convolutional neural network image recognition program. DETAILED DESCRIPTION
[0018] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the relevant invention are shown in the accompanying drawings.
[0019] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.
[0020] The present invention provides a linear engineering slope disease detection method, the method comprising: Collecting multivariate contour information of the target area; the multivariate contour information includes aerial images, depth images and laser detection point clouds; Aligning the multivariate contour information, and setting the multivariate contour information category weights according to the aerial image; Construct a three-dimensional digital geological model of the slope based on weighted multivariate contours; Extracting geometric parameters and geological parameters of a designated two-dimensional slope surface according to the three-dimensional digital geological model of the slope; Based on the geometric parameters and geological parameters, a slope safety factor F is calculated as a slope safety factor of a specified two-dimensional slope surface, and the stability of the slope is evaluated by the slope safety factor; The slope safety factor F is: ; represents the safety factor; It indicates the soil bulk density; Indicates adhesion; represents the friction angle; Indicates slope height; represents the downhill angle; From the slope uphill angle Determined undetermined constants; The slope downslope angle Determine the undetermined constant.
[0021] In order to more clearly explain the linear engineering slope disease detection method, system and equipment of the present invention, Figure 1 , Figure 2 , Figure 3Each step in the embodiment of the present invention is described in detail.
[0022] A linear engineering slope disease detection method according to an embodiment of the present invention includes: Step S1, collecting multivariate contour information of the target area; the multivariate contour information includes aerial images, depth images and laser detection point clouds.
[0023] In this embodiment, the method for obtaining the multivariate profile information is: The information collection equipment continuously collects the slope profile information of each dimension in the target area according to the respective set routes; The slope profile information of different dimensions is aligned and / or data fused through preset markers or automatically identified key points to obtain multivariate profile information.
[0024] The information collection equipment can be mounted on the drone 6, and the image recognition program 7 can be embedded in the drone to serve as the slope disease intelligent recognition module 1. Multiple contour information can be collected simultaneously, at multiple points, and in multiple dimensions through one or more slope disease intelligent recognition modules. The drone 6 automatically sends the collected information to the image recognition program 7. The multiple and multiple point information can be aligned and matched by adding marking points to the collected contour information or by setting a specific category of marker images.
[0025] Step S2, aligning the multivariate contour information and setting the multivariate contour information category weights according to the aerial image; Step S3, constructing a three-dimensional digital geological model of the slope according to the weighted multivariate contours; In this embodiment, the construction of the three-dimensional digital geological model of the slope includes: Based on the characteristic parameters of multivariate contour information, the three-dimensional reconstruction model of the slope is obtained through the preset image matching and motion recovery structure algorithm; The three-dimensional reconstructed model of the slope is converted into a point cloud data model, and the point cloud data model is densely reconstructed based on a depth map fusion method to obtain an elevation cloud map model of the slope; According to the position information of the dense points in the elevation cloud model, the elevation cloud model is discretized to obtain a three-dimensional digital model of the diseased slope at the site scale.
[0026] The discretization process can be performed by triangulating the surface of the three-dimensional reconstructed model of the diseased slope to obtain the three-dimensional terrain of the diseased slope at the site scale, and the properties of the slope rock and soil, the rock-soil interface, and the statistical information of joints and fissures can be identified based on high-definition images and convolutional neural network algorithms to establish a three-dimensional digital geological model of the diseased slope. The process of constructing the three-dimensional digital geological model of the diseased slope is carried out on the indoor server platform 5, which includes an image information transmission and acquisition module 4, a three-dimensional model reconstruction module 2 of the diseased slope, and a rapid assessment model 3 for the stability of the diseased slope.
[0027] Step S4, extracting geometric parameters and geological parameters of a designated two-dimensional slope surface according to the three-dimensional digital geological model of the slope; In this embodiment, the specified two-dimensional slope is selected by: Extracting slope image features based on the aerial image; Based on the slope image features, a slope disease recognition result is obtained through a disease recognition network based on a convolutional neural network; That is, this embodiment essentially first identifies the slopes that may have slope diseases, and then conducts targeted analysis on the selected slopes. The process of the linear engineering slope disease intelligent assessment method with a specified two-dimensional slope is as follows: Figure 4 shown.
[0028] In this embodiment, the training method of the disease recognition network based on the convolutional neural network includes: Perform feature extraction and cluster analysis on the labeled aerial images to obtain the feature parameters of the historical diseased slope images and the feature parameters of the historical disease-free slope images; In this embodiment, feature extraction and cluster analysis are performed on the aerial images with labels: Determining different characteristic parameter combinations based on the historical aerial images in the initial database of slope diseases; Clustering the historical aerial images according to the different characteristic parameter combinations to obtain clustering results; According to the clustering result, characteristic parameters of historical diseased slope images are screened out from historical aerial images.
[0029] Historical aerial images can be obtained through network digital crawler technology, and combined with existing on-site slope disease images to establish a massive initial database of slope diseases. A cluster analysis method is used to establish a classification system for slope disease vector diagrams, and "training-testing" of slope diseases is carried out based on deep convolutional neural networks.
[0030] The characteristics of the historical diseased slope images include landslide, rockfall, collapse, crack, protective structure break, protective structure fall-off, and no disease, a total of 7. The image recognition program 7 selected is an image recognition convolutional neural network with self-learning function.
[0031] The characteristic parameters of the diseased slope image and the characteristic parameters of the historical disease-free slope image are mixed and passed through the disease recognition network based on the convolutional neural network to be trained to obtain the slope disease recognition result of the training set, the model parameters are adjusted and iterated to the preset termination condition, and the trained disease recognition network based on the convolutional neural network is obtained.
[0032] The two-dimensional slope surface with slope disease identification results indicating that it has disease is selected as the designated two-dimensional slope surface.
[0033] Step S5, calculating the slope safety factor based on the geometric parameters and geological parameters As a slope safety factor of a specified two-dimensional slope surface, the stability of the slope is evaluated by the slope safety factor; The slope safety factor for: ; represents the safety factor; It indicates the soil bulk density; Indicates adhesion; represents the friction angle; Indicates slope height; represents the downhill angle; From the slope uphill angle Determined undetermined constants; The slope downslope angle Determine the undetermined constant.
[0034] The relationship between the upslope angle and the downslope angle is shown in Table 1: Table 1 Values of undetermined parameters
[0035] The second embodiment of the present invention discloses a linear engineering slope disease detection system, the system comprising: A multivariate contour information acquisition module is used to collect multivariate contour information of a target area through different sensors; the multivariate contour information includes an aerial image, a depth image, and a laser detection point cloud; A data alignment module, used for aligning the multivariate profile information and setting the multivariate profile information category weights according to the aerial image; The diseased slope 3D model reconstruction module is used to construct a 3D digital geological model of the slope based on weighted multivariate contours; A parameter extraction module, used for extracting geometric parameters and geological parameters of a designated two-dimensional slope surface according to the three-dimensional digital geological model of the slope; Diseased slope stability assessment module, used for Based on the geometric and geological parameters, calculate the slope safety factor As a slope safety factor of a specified two-dimensional slope surface, the stability of the slope is evaluated by the slope safety factor; The slope safety factor for: ; represents the safety factor; It indicates the soil bulk density; Indicates adhesion; represents the friction angle; Indicates slope height; represents the downhill angle; From the slope uphill angle Determined undetermined constants; The slope downslope angle Determine the undetermined constant.
[0036] A device according to a third embodiment of the present invention includes: at least one processor; and a memory communicatively connected to at least one of the processors; wherein, The memory stores instructions that can be executed by the processor, and the instructions are used to be executed by the processor to implement a linear engineering slope disease detection method.
[0037] A fourth embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to be executed by the computer to implement a linear engineering slope disease detection method.
[0038] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related instructions of the storage device and processing device described above can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.
[0039] In the description of the present invention, terms such as "center", "up", "down", "left", "right", "vertical", "horizontal", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be understood as a limitation on the present invention.
[0040] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0041] The term "comprise" or any other similar term is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus / device that includes a list of elements includes not only those elements but also other elements not expressly listed, or also includes elements inherent to such process, method, article, or apparatus / device.
[0042] 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 linear engineering slope disease detection method, characterized in that: The linear engineering slope disease detection method comprises: Collecting multivariate contour information of the target area; the multivariate contour information includes aerial images, depth images and laser detection point clouds; Aligning the multivariate contour information and setting the multivariate contour information category weights according to the aerial image; Construct a three-dimensional digital geological model of the slope based on weighted multivariate contours; Extracting geometric parameters and geological parameters of a designated two-dimensional slope surface according to the three-dimensional digital geological model of the slope; Based on the geometric and geological parameters, calculate the slope safety factor As a slope safety factor of a specified two-dimensional slope surface, the stability of the slope is evaluated by the slope safety factor; The slope safety factor for: ; Where, represents the safety factor; Indicates soil bulk density; Indicates adhesion; represents the friction angle; Indicates slope height; Indicates the downhill angle; 、 All are determined by the upslope angle of the slope and downhill angle Determine the undetermined constant.
2. A linear engineering slope disease detection method according to claim 1, characterized in that: The method for obtaining the multivariate profile information is as follows: The information collection equipment continuously collects slope profile information of various dimensions in the target area according to their respective set routes; The slope profile information of different dimensions is aligned and / or data fused through preset markers or automatically identified key points to obtain multivariate profile information.
3. A linear engineering slope disease detection method according to claim 1, characterized in that: The method for selecting the specified two-dimensional slope is as follows: extracting slope image features based on the aerial image; Based on the slope image features, a slope disease recognition result is obtained through a disease recognition network based on a convolutional neural network; The two-dimensional slope surface with slope disease identification results indicating that it has diseases is selected as the designated two-dimensional slope surface.
4. A linear engineering slope disease detection method according to claim 3, characterized in that: The training method of the disease recognition network based on convolutional neural network includes: Perform feature extraction and cluster analysis on labeled aerial images to obtain the characteristic parameters of historical diseased slope images and historical disease-free slope images; The characteristic parameters of the diseased slope image and the characteristic parameters of the historical disease-free slope image are mixed and passed through the disease recognition network based on the convolutional neural network to be trained to obtain the slope disease recognition results of the training set. The model parameters are adjusted and iterated to the preset termination conditions to obtain the trained disease recognition network based on the convolutional neural network.
5. A linear engineering slope disease detection method according to claim 4, characterized in that: The construction of the three-dimensional digital geological model of the slope includes: Based on the characteristic parameters of the multivariate contour information, a 3D reconstructed model of the slope is obtained through a preset image matching and structure-from-motion algorithm. Converting the three-dimensional reconstructed model of the slope into a point cloud data model, and densely reconstructing the point cloud data model based on a depth map fusion method to obtain an elevation cloud model of the slope; According to the position information of the dense points in the elevation cloud model, the elevation cloud model is discretized to obtain a three-dimensional digital model of the diseased slope at the site scale.
6. A linear engineering slope disease detection method according to claim 5, characterized in that: Perform feature extraction and cluster analysis on labeled aerial images: determining different characteristic parameter combinations based on historical aerial images in the initial database of slope diseases; Clustering the historical aerial images according to the different feature parameter combinations to obtain clustering results; According to the clustering result, characteristic parameters of historical diseased slope images are screened from historical aerial images.
7. A linear engineering slope disease detection system, characterized in that: The system comprises: A multi-dimensional contour information acquisition module is used to collect multi-dimensional contour information of the target area through different sensors; the multi-dimensional contour information includes aerial images, depth images and laser detection point clouds; a data alignment module, configured to align the multivariate profile information and set the multivariate profile information category weights according to the aerial image; The diseased slope 3D model reconstruction module is used to construct a 3D digital geological model of the slope based on weighted multivariate contours; A parameter extraction module, configured to extract geometric parameters and geological parameters of a designated two-dimensional slope surface based on the three-dimensional digital geological model of the slope; Diseased slope stability assessment module, used for Based on the geometric and geological parameters, calculate the slope safety factor As a slope safety factor of a specified two-dimensional slope surface, the stability of the slope is evaluated by the slope safety factor; The slope safety factor for: ; Where, represents the safety factor; Indicates soil bulk density; Indicates adhesion; represents the friction angle; Indicates slope height; Indicates the downhill angle; 、 All are determined by the upslope angle of the slope and downhill angle Determine the undetermined constant.
8. A linear engineering slope disease detection system according to claim 7, characterized in that: The diseased slope three-dimensional model reconstruction module, the diseased slope stability rapid assessment module and the image information transmission and acquisition module are integrated on an indoor server platform.
9. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively connected to at least one of the processors; wherein, The memory stores instructions that can be executed by the processor, and the instructions are used to be executed by the processor to implement the linear engineering slope disease detection method according to any one of claims 1 to 6.
10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to be executed by the computer to implement the linear engineering slope disease detection method according to any one of claims 1 to 6.
Citation Information
Cited By
Protection device design method and system based on BIM technology
CN120372888A
