Roadway displacement field detection method and device
By constructing point cloud hierarchical features and layer-by-layer inversion technology, the high cost problem of roadway surrounding rock deformation monitoring system was solved, high-precision detection of displacement across the entire field and detailed stability assessment were achieved, the cost of sensor deployment was reduced, and the efficiency of roadway excavation was improved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-19
- Publication Date
- 2026-04-14
AI Technical Summary
Existing roadway surrounding rock deformation monitoring systems are costly to deploy and maintain, making it difficult to monitor and assess the deformation field across the entire area. Furthermore, three-dimensional laser scanning technology cannot provide accurate point-to-point matching relationships between multi-stage point clouds, resulting in calculation results that are not applicable to displacement field calculations.
By constructing point cloud hierarchical features and performing layer-by-layer inversion, including point-to-point matching at the anchor tray level and metal mesh level, the inverse distance weighted interpolation method is used to determine the full-field displacement vector and generate the displacement field of the tunnel area.
It achieves high-precision inversion of displacement across the entire field, reduces detection costs, simplifies construction processes, improves tunnel excavation efficiency, and provides more detailed and accurate tunnel stability assessment data.
Smart Images
Figure CN121855439A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of coal mine roadway detection technology, and in particular to a method and apparatus for detecting roadway displacement field. Background Technology
[0002] A key aspect of roadway surrounding rock stability assessment is the long-term monitoring of surrounding rock deformation. Traditional monitoring methods include visual deformation inspection, manual measurement, and more advanced technologies such as roof and floor convergence monitors and cross-measuring lines. With the development of sensor technology, automated measurement tools provide support systems with richer data. However, the problem of comprehensively assessing roadway surrounding rock deformation remains unresolved, mainly because the deployment and maintenance costs of existing monitoring systems are high, resulting in these systems typically being limited to key nodes and areas, making it difficult to monitor and assess the entire deformation field.
[0003] The non-destructive testing and reusable equipment of 3D laser scanning technology give it a significant cost advantage. However, the full-field displacement inversion technology for roadway surrounding rock based on 3D laser scanning cannot provide accurate point-to-point matching relationships between multi-stage point clouds, which means that the calculation results are not applicable to displacement field calculations. Summary of the Invention
[0004] This disclosure aims to at least partially address one of the technical problems in the related art.
[0005] Therefore, the first aspect of this disclosure proposes a method for detecting displacement fields in roadways, comprising the following steps: Acquire source point cloud data and deformed point cloud data of the target tunnel area, wherein the target tunnel area includes anchor trays and metal mesh; Point cloud hierarchical features are constructed for the source point cloud data and the deformed point cloud data respectively, to obtain the first anchor tray hierarchical point cloud and the first metal mesh hierarchical point cloud in the source point cloud data, and the second anchor tray hierarchical point cloud and the second metal mesh hierarchical point cloud in the deformed point cloud data. After performing point cloud registration on the source point cloud data and the deformed point cloud data, point-to-point matching at the anchor tray level is performed on the first anchor tray level point cloud and the second anchor tray level point cloud. Based on the point-to-point matching result at the anchor tray level, the first displacement vector of the first anchor tray level point cloud is determined. Based on the first displacement vector, point-to-point matching at the metal mesh level is performed on the first metal mesh level point cloud and the second metal mesh level point cloud, and the second displacement vector of the first metal mesh level point cloud is determined based on the point-to-point matching result at the metal mesh level. Using the first displacement vector and the second displacement vector, the third displacement vector of the point cloud other than the first anchor tray level point cloud and the first metal mesh level point cloud in the source point cloud data is determined by inverse distance weighted interpolation.
[0006] A second aspect of this disclosure provides a tunnel displacement field detection device, comprising: The acquisition module is used to acquire source point cloud data and deformed point cloud data of the target tunnel area, wherein the target tunnel area includes anchor trays and metal mesh; The hierarchical feature extraction module is used to construct the hierarchical features of the source point cloud data and the deformed point cloud data respectively, and obtain the first anchor tray hierarchical point cloud and the first metal mesh hierarchical point cloud in the source point cloud data, and the second anchor tray hierarchical point cloud and the second metal mesh hierarchical point cloud in the deformed point cloud data. An anchor tray level matching module is used to perform point cloud registration on the source point cloud data and the deformed point cloud data, and then perform point-to-point matching on the first anchor tray level point cloud and the second anchor tray level point cloud, and determine the first displacement vector of the first anchor tray level point cloud based on the point-to-point matching result of the anchor tray level. The metal mesh layer matching module is used to perform metal mesh layer point pair matching on the first metal mesh layer point cloud and the second metal mesh layer point cloud based on the first displacement vector, and to determine the second displacement vector of the first metal mesh layer point cloud based on the metal mesh layer point pair matching result. Other hierarchical matching modules are used to determine the third displacement vector of the point cloud other than the first anchor tray level point cloud and the first metal mesh level point cloud in the source point cloud data by using the first displacement vector and the second displacement vector through inverse distance weighted interpolation.
[0007] A third aspect of this disclosure provides an electronic device, including: a processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method described in the first aspect above.
[0008] A fourth aspect of this disclosure provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method described in the first aspect above.
[0009] The roadway displacement field detection method disclosed herein obtains displacement vectors corresponding to point clouds at different levels through layer-by-layer inversion. This includes the first displacement vector of the first anchor tray level point cloud, the second displacement vector of the first metal mesh level point cloud, and the third displacement vector of other level point clouds. This covers all point cloud data of the target roadway area, generating a displacement field for the target roadway area. This method comprehensively captures the displacement distribution of the roadway cross-section and improves displacement detection accuracy, overcoming the limitations of traditional methods that can only obtain single-point or local displacements. It achieves high-precision inversion of the entire displacement field of coal mine roadways, thus providing more detailed and accurate data support for roadway stability assessment. Furthermore, this embodiment enables low-cost, high-precision detection of roadway displacement without requiring complex manual labor or expensive detection equipment. This significantly reduces the cost of sensor deployment, simplifies the construction process, and improves roadway excavation efficiency.
[0010] Additional aspects and advantages of this disclosure will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this disclosure. Attached Figure Description
[0011] The above and / or additional aspects and advantages of this disclosure will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, in which: Figure 1 A schematic flowchart of a roadway displacement field detection method provided in an embodiment of this disclosure; Figure 2 A schematic diagram of a processed point cloud hierarchical feature provided in an embodiment of this disclosure; Figure 3 A schematic diagram of a hierarchical point cloud of an anchor tray and a metal mesh after hierarchical extraction, provided for embodiments of this disclosure; Figure 4 A schematic diagram of registered source point cloud data and deformed point cloud data provided in an embodiment of this disclosure; Figure 5 This is a schematic diagram illustrating the aggregation processing of point clouds at the anchor tray level, provided by an embodiment of the present disclosure. Figure 6 A schematic diagram of a metal mesh layered point cloud provided in an embodiment of this disclosure; Figure 7 An experimental prediction error distribution diagram provided for embodiments of this disclosure; Figure 8 This is a schematic diagram of a tunnel displacement field detection device provided in an embodiment of this disclosure. Detailed Implementation
[0012] Embodiments of this disclosure are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this disclosure, and should not be construed as limiting this disclosure.
[0013] Specifically, the method and apparatus for detecting roadway displacement field according to embodiments of the present disclosure are described below with reference to the accompanying drawings.
[0014] Figure 1 This is a schematic flowchart illustrating a method for detecting displacement fields in a roadway, as provided in an embodiment of this disclosure. Figure 1 As shown, the tunnel displacement field detection method may include the following steps: Step 101: Obtain source point cloud data and deformed point cloud data of the target tunnel area, which includes anchor bolt trays and metal mesh.
[0015] The source point cloud data and the deformed point cloud data are point cloud data of the target tunnel area at different stages. Optionally, the target tunnel area can be determined based on the actual support design and geological conditions; for example, the target tunnel area can be the roof area, or it can be the tunnel sidewalls or floor. Step 102: Construct point cloud hierarchical features for the source point cloud data and the deformed point cloud data respectively, to obtain the first anchor tray level point cloud and the first metal mesh level point cloud in the source point cloud data, and the second anchor tray level point cloud and the second metal mesh level point cloud in the deformed point cloud data.
[0016] It should be noted that multi-stage point cloud registration algorithms in related technologies suffer from invariance of registration markers and high sensitivity to the initial registration matrix. For example, pre-placing target spheres for registration of point clouds at different stages leads to a sharp increase in workload. Furthermore, small-scale local target spheres cannot guarantee sufficient accuracy for large-scale point cloud registration. Since the point cloud data obtained from tunnel scanning contains a massive number of redundant points, interfering with subsequent point cloud registration, this embodiment of the present disclosure constructs point cloud hierarchical features to separate the main components, namely the anchor bolt tray and the metal mesh (metal tube), as key features for global registration and point pair matching.
[0017] In one implementation, the point cloud hierarchical features of the source point cloud data and the deformed point cloud data can be constructed through the following steps: S1. Establish a KDTree spatial index and use a hybrid search strategy to determine the neighborhood point set of each point in the source point cloud data and the deformed point cloud data. ;
[0018] in, The coordinates of the target point, The maximum search radius (exemplarily, 0.002m) is used. This represents the maximum number of neighboring points (for example, a value between 30 and 50).
[0019] S2, calculate the corresponding covariance matrix based on the neighborhood point set of each point in the source point cloud data and the deformed point cloud data. ;
[0020] S3. Perform spectral decomposition on the covariance matrix to find its minimum eigenvalue. This yields the hierarchical features of each point in the source point cloud data and the deformed point cloud data. S4. Based on the hierarchical features, determine the first anchor tray level point cloud and the first metal mesh level point cloud in the source point cloud data, as well as the second anchor tray level point cloud and the second metal mesh level point cloud in the deformed point cloud data.
[0021] Among them, the numerical range of the hierarchical feature λ differs for different main components. Figure 2 This is a schematic diagram of a processed point cloud hierarchical feature provided in an embodiment of this disclosure. For example... Figure 2 As shown, different colors represent different feature values. Therefore, the anchor tray level point cloud, metal mesh level point cloud, and other level point clouds (such as surrounding rock) in the point cloud data can be separated based on the preset feature value ranges constructed by different subjects. Figure 3 This is a schematic diagram of a hierarchical point cloud of an anchor tray and a metal mesh after hierarchical extraction, provided in an embodiment of this disclosure.
[0022] Step 103: After registering the source point cloud data and the deformed point cloud data, perform point-to-point matching of the first anchor tray level point cloud and the second anchor tray level point cloud at the anchor tray level. Based on the point-to-point matching result at the anchor tray level, determine the first displacement vector of the first anchor tray level point cloud.
[0023] In some embodiments of this disclosure, to address the problem of requiring a high-precision initial transformation matrix for redundant point cloud registration, a method for point cloud registration based on non-redundant point cloud data is proposed, as shown in the following steps: S1, using the point cloud of the first anchor tray layer and the point cloud of the first metal mesh layer as non-redundant source point cloud data, and the point cloud of the second anchor tray layer and the point cloud of the second metal mesh layer as non-redundant deformable point cloud data, the RANSAC algorithm is used to perform coarse registration on the non-redundant source point cloud data and the non-redundant deformable point cloud data to obtain the initial estimated transformation matrix, including the rotation matrix. Translation matrix ; S2, Based on the initial estimated transformation matrix, the non-redundant deformed point cloud data is processed to obtain the initial estimated deformed point cloud;
[0024] in, For deformed point cloud data (at the anchor tray level or metal mesh level), For the initial estimation of deformed point cloud S3. The Iterative Closest Point (ICP) algorithm is used to perform fine registration of the non-redundant source point cloud data and the initially estimated deformed point cloud, resulting in a fine registration transformation matrix. and ; S4. Based on the fine registration transformation matrix, the initially estimated deformed point cloud is processed to obtain the registered deformed point cloud data.
[0025]
[0026] in, The deformed point cloud data after registration. Figure 4 This is a schematic diagram of registered source point cloud data and deformed point cloud data provided in an embodiment of this disclosure. Figure 4 Medium-dark gray represents the source point cloud, and light gray represents the deformed point cloud. By making full use of the numerous multi-scale features naturally present on the surface of coal mine roadways, a two-step registration strategy combining coarse and fine registration is adopted, which significantly improves the registration accuracy and efficiency of large-scale point clouds.
[0027] Optionally, in some embodiments, the point cloud normal vectors can also be modified: S1, Normal Vector Calculation: For each point Based on neighborhood point set Calculate the covariance matrix and find the eigenvector corresponding to the smallest eigenvalue of the covariance matrix. As the initial normal vector.
[0028] in The coordinates of the neighborhood centroid are given.
[0029] S2, Normal Vector Direction Correction: Defines the overall centroid of the tunnel. Force the normal vector to point inside the tunnel:
[0030] In some embodiments of this disclosure, to address the difficulty in obtaining point-pair matching relationships, a layer-by-layer inversion algorithm is proposed to perform high-precision full-field displacement field evaluation based on hierarchical features. The following steps can be referred to: S1, perform aggregation processing on the first anchor tray level point cloud and the second anchor tray level point cloud respectively to obtain the first anchor tray aggregation point corresponding to the first anchor tray level point cloud and the second anchor tray aggregation point corresponding to the second anchor tray level point cloud. Pallets are an essential component in tunnel support. The size of a pallet is approximately 0.15m x 0.15m, and it can be considered a rigid component. Therefore, the centroid of pallet clustering can be used as one of the corresponding point pairs. Figure 5 This is a schematic diagram illustrating the aggregation processing of point clouds at the anchor bolt tray level, provided as an embodiment of this disclosure. Figure 5 As shown, the gray area represents the point cloud of the anchor tray hierarchy, and the black area represents the aggregation points of the anchor tray.
[0031] S2, the best matching point pair between the first anchor tray aggregation point and the second anchor tray aggregation point is obtained by the nearest point filtering; S3, based on the displacement vector of the best matching point pair between the first anchor tray aggregation point and the second anchor tray aggregation point, the first displacement vector of the first anchor tray hierarchical point cloud (including aggregation points and non-aggregation points) is determined by interpolation.
[0032] Step 104: Based on the first displacement vector, perform point-to-point matching at the metal mesh level on the first metal mesh layer point cloud and the second metal mesh layer point cloud, and determine the second displacement vector of the first metal mesh layer point cloud based on the point-to-point matching result at the metal mesh level.
[0033] Metal mesh is the second obvious hierarchical feature of point cloud obtained by tunnel scanning. It is mainly a rectangular grid metal mesh with a diameter of about 0.01m, so it can be screened through the second-level feature. Figure 6 This is a schematic diagram of a metal mesh layered point cloud provided in an embodiment of this disclosure. Figure 6 As shown, the features between metal meshes are not obvious. Therefore, in this embodiment, the correspondence between the point clouds of the first metal mesh layer and the point clouds of the second metal mesh layer is determined by the first displacement vector of the anchor tray layer, and a method for screening candidate domains is proposed: S1, Deform the first metal mesh layer point cloud based on the first displacement vector to obtain the initial metal mesh deformation point corresponding to each point in the first metal mesh layer point cloud;
[0034] in, Source Point Cloud The i-th point in Let be the first displacement vector. This is the initial metal mesh deformation point corresponding to the i-th point in the first metal mesh layer point cloud.
[0035] S2, Based on the coordinates of the initial deformed point cloud of the metal mesh, determine the candidate neighborhood point set for metal mesh layer point pair matching corresponding to each initial deformed point cloud in the second metal mesh layer point cloud.
[0036] in, For deformable point clouds Corresponding candidate neighborhood S3, based on the first metal mesh layer point cloud and the candidate neighborhood point set corresponding to each point in the first metal mesh layer point cloud for metal mesh layer point pair matching, the best matching point pair between the first metal mesh layer point cloud and the second metal mesh layer point cloud is obtained by the nearest point filtering;
[0037] in, The representative selects the distance in the candidate neighborhood. The nearest point, Representative deformable point cloud neutralization The matching point cloud.
[0038] S4. Determine the second displacement vector of the first metal mesh layer point cloud based on the best matching point pair between the first metal mesh layer point cloud and the second metal mesh layer point cloud.
[0039] Step 105: Using the first displacement vector and the second displacement vector, determine the third displacement vector of the point cloud other than the first anchor tray level point cloud and the first metal mesh level point cloud in the source point cloud data by inverse distance weighted interpolation.
[0040] Alternatively, the third displacement vector of an unknown point in the source point cloud data can be determined using the following formula:
[0041] in, Represents the known set of points (Point cloud of the first anchor tray level and point cloud of the first metal mesh level) with target point The set of points in the neighborhood of the center, represent The i-th point in the middle, This represents its corresponding displacement. This represents the weight of that point. This represents the distance from the target point to this point. This represents the displacement vector to be determined at the target point.
[0042] By implementing the embodiments of this disclosure, displacement vectors corresponding to point clouds at different levels are obtained through layer-by-layer inversion, including the first displacement vector of the first anchor tray level point cloud, the second displacement vector of the first metal mesh level point cloud, and the third displacement vector of other level point clouds. This covers all point cloud data of the target roadway area, and based on this, a displacement field of the target roadway area is generated. This comprehensively captures the displacement distribution of the roadway cross-section and improves the displacement detection accuracy, overcoming the limitations of traditional methods that can only obtain single-point or local displacements. It achieves high-precision inversion of the entire displacement field of coal mine roadways, thus providing more detailed and accurate data support for roadway stability assessment. In addition, this embodiment can also achieve low-cost and high-precision detection of roadway displacement without the need for complicated manual labor and expensive detection equipment. This can greatly reduce the cost problems caused by sensor deployment, simplify the construction process, and improve the efficiency of roadway excavation.
[0043] The method for measuring the deformation of the entire roadway in this embodiment provides more complete data compared to traditional single-point measurement methods, as the deformation response at any point in the roadway can be obtained. This means that roadway engineers can analyze and monitor any area of the roadway to identify risk zones and take appropriate safety measures. In contrast, traditional methods can only provide deformation records for local measurement points and cannot represent the overall operating status of the roadway.
[0044] Figure 7 This is a distribution diagram of experimental prediction errors provided for embodiments of this disclosure. (For example...) Figure 7 As shown in the figure, experimental verification shows that the displacement field inversion accuracy of this framework reaches the millimeter level, which is significantly better than traditional displacement field detection technology.
[0045] Figure 8 This is a schematic diagram of a roadway displacement field detection device provided in an embodiment of this disclosure. Figure 8 As shown, the tunnel displacement field detection device may include: an acquisition module 801, a hierarchical feature extraction module 802, an anchor bolt tray hierarchical matching module 803, a metal mesh hierarchical matching module 804, and other hierarchical matching modules 805.
[0046] The acquisition module 801 is used to acquire source point cloud data and deformed point cloud data of the target tunnel area, which includes anchor trays and metal mesh.
[0047] The hierarchical feature extraction module 802 is used to construct the point cloud hierarchical features of the source point cloud data and the deformed point cloud data respectively, to obtain the first anchor tray hierarchical point cloud and the first metal mesh hierarchical point cloud in the source point cloud data, and the second anchor tray hierarchical point cloud and the second metal mesh hierarchical point cloud in the deformed point cloud data.
[0048] The anchor tray level matching module 803 is used to perform point cloud registration on the source point cloud data and the deformed point cloud data, and then perform point-to-point matching of the first anchor tray level point cloud and the second anchor tray level point cloud at the anchor tray level. Based on the point-to-point matching result of the anchor tray level, the first displacement vector of the first anchor tray level point cloud is determined.
[0049] The metal mesh layer matching module 804 is used to perform metal mesh layer point pair matching on the first metal mesh layer point cloud and the second metal mesh layer point cloud based on the first displacement vector, and to determine the second displacement vector of the first metal mesh layer point cloud based on the metal mesh layer point pair matching result.
[0050] Other level matching modules 805 are used to determine the third displacement vector of the point cloud other than the first anchor tray level point cloud and the first metal mesh level point cloud in the source point cloud data by using the first displacement vector and the second displacement vector through inverse distance weight interpolation.
[0051] In some embodiments of this disclosure, the hierarchical feature construction module 802 is specifically used for: establishing a KDTree spatial index; using a hybrid search strategy to determine the neighborhood point set of each point in the source point cloud data and the deformed point cloud data; calculating the corresponding covariance matrix based on the neighborhood point set of each point in the source point cloud data and the deformed point cloud data; performing spectral decomposition on the covariance matrix to obtain the hierarchical features of each point in the source point cloud data and the deformed point cloud data; and determining the first anchor tray hierarchical point cloud and the first metal mesh hierarchical point cloud in the source point cloud data, as well as the second anchor tray hierarchical point cloud and the second metal mesh hierarchical point cloud in the deformed point cloud data based on the hierarchical features.
[0052] In some embodiments of this disclosure, the anchor tray level matching module 803 is specifically used to: perform aggregation processing on the first anchor tray level point cloud and the second anchor tray level point cloud respectively to obtain a first anchor tray aggregation point corresponding to the first anchor tray level point cloud and a second anchor tray aggregation point corresponding to the second anchor tray level point cloud; obtain the best matching point pair between the first anchor tray aggregation point and the second anchor tray aggregation point through nearest point filtering; and determine the first displacement vector of the first anchor tray level point cloud by interpolation based on the displacement vector of the best matching point pair between the first anchor tray aggregation point and the second anchor tray aggregation point.
[0053] In some embodiments of this disclosure, the metal mesh layer matching module 804 is specifically used for: deforming the first metal mesh layer point cloud based on the first displacement vector to obtain an initial metal mesh deformed point corresponding to each point in the first metal mesh layer point cloud; determining a candidate neighborhood point set for metal mesh layer point pair matching for each initial metal mesh deformed point cloud in the second metal mesh layer point cloud based on the coordinates of the initial metal mesh deformed point cloud; obtaining the best matching point pair between the first metal mesh layer point cloud and the second metal mesh layer point cloud by nearest point filtering based on the first metal mesh layer point cloud and the candidate neighborhood point set for metal mesh layer point pair matching for each point in the first metal mesh layer point cloud; and determining the second displacement vector of the first metal mesh layer point cloud based on the best matching point pair between the first metal mesh layer point cloud and the second metal mesh layer point cloud.
[0054] In some embodiments of this disclosure, such as Figure 8 Based on the illustrated embodiment, the tunnel displacement field detection device may further include a point cloud registration module. This point cloud registration module can be used to: use the first anchor tray layer point cloud and the first metal mesh layer point cloud as non-redundant source point cloud data, and the second anchor tray layer point cloud and the second metal mesh layer point cloud as non-redundant deformed point cloud data; perform coarse registration on the non-redundant source point cloud data and the non-redundant deformed point cloud data using the RANSAC algorithm to obtain an initial estimated transformation matrix; process the non-redundant deformed point cloud data based on the initial estimated transformation matrix to obtain an initial estimated deformed point cloud; perform fine registration on the non-redundant source point cloud data and the initial estimated deformed point cloud using the Iterative Closest Point Algorithm (ICP) to obtain a fine registration transformation matrix; and process the initial estimated deformed point cloud based on the fine registration transformation matrix to obtain registered deformed point cloud data.
[0055] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.
[0056] To implement the above embodiments, this disclosure also proposes an electronic device, including: a processor and a memory communicatively connected to the processor; the memory stores computer execution instructions; the processor executes the computer execution instructions stored in the memory to implement the method provided in the foregoing embodiments.
[0057] To implement the above embodiments, this disclosure also proposes a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the methods provided in the foregoing embodiments.
[0058] To implement the above embodiments, this disclosure also proposes a computer program product, including a computer program that, when executed by a processor, implements the methods provided in the foregoing embodiments.
[0059] In the foregoing descriptions of the embodiments, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.
[0060] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this disclosure, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.
[0061] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of this disclosure includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of this disclosure pertain.
[0062] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.
[0063] It should be understood that various parts of this disclosure can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0064] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it includes one or a combination of the steps of the method embodiments.
[0065] Furthermore, the functional units in the various embodiments of this disclosure can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.
[0066] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present disclosure have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present disclosure. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present disclosure.
Claims
1. A method for detecting displacement field in roadways, characterized in that, Includes the following steps: Acquire source point cloud data and deformed point cloud data of the target tunnel area, wherein the target tunnel area includes anchor trays and metal mesh; Point cloud hierarchical features are constructed for the source point cloud data and the deformed point cloud data respectively, to obtain the first anchor tray hierarchical point cloud and the first metal mesh hierarchical point cloud in the source point cloud data, and the second anchor tray hierarchical point cloud and the second metal mesh hierarchical point cloud in the deformed point cloud data. After performing point cloud registration on the source point cloud data and the deformed point cloud data, point-to-point matching at the anchor tray level is performed on the first anchor tray level point cloud and the second anchor tray level point cloud. Based on the point-to-point matching result at the anchor tray level, the first displacement vector of the first anchor tray level point cloud is determined. Based on the first displacement vector, point-to-point matching at the metal mesh level is performed on the first metal mesh level point cloud and the second metal mesh level point cloud, and the second displacement vector of the first metal mesh level point cloud is determined based on the point-to-point matching result at the metal mesh level. Using the first displacement vector and the second displacement vector, the third displacement vector of the point cloud other than the first anchor tray level point cloud and the first metal mesh level point cloud in the source point cloud data is determined by inverse distance weighted interpolation.
2. The method according to claim 1, characterized in that, The step of constructing point cloud hierarchical features for the source point cloud data and the deformed point cloud data respectively, to obtain the first anchor tray level point cloud and the first metal mesh level point cloud in the source point cloud data, and the second anchor tray level point cloud and the second metal mesh level point cloud in the deformed point cloud data, includes: A KDTree spatial index is established, and a hybrid search strategy is used to determine the neighborhood point set of each point in the source point cloud data and the deformed point cloud data. The covariance matrix is calculated based on the neighborhood point set of each point in the source point cloud data and the deformed point cloud data. Spectral decomposition is performed on the covariance matrix to obtain the hierarchical features of each point in the source point cloud data and the deformed point cloud data; Based on the hierarchical features, the first anchor tray hierarchical point cloud and the first metal mesh hierarchical point cloud in the source point cloud data, as well as the second anchor tray hierarchical point cloud and the second metal mesh hierarchical point cloud in the deformed point cloud data, are determined.
3. The method according to claim 1, characterized in that, The step of performing point-to-point matching at the anchor tray level on the first anchor tray level point cloud and the second anchor tray level point cloud, and determining the first displacement vector of the first anchor tray level point cloud based on the point-to-point matching result at the anchor tray level, includes: The first anchor tray level point cloud and the second anchor tray level point cloud are respectively aggregated to obtain the first anchor tray aggregation point corresponding to the first anchor tray level point cloud and the second anchor tray aggregation point corresponding to the second anchor tray level point cloud. The best matching point pair between the first anchor tray aggregation point and the second anchor tray aggregation point is obtained by the nearest point filtering; The first displacement vector of the first anchor tray level point cloud is determined by interpolation based on the displacement vector of the best matching point pair between the first anchor tray aggregation point and the second anchor tray aggregation point.
4. The method according to claim 1, characterized in that, The step of performing point-to-point matching at the metal mesh level on the first and second metal mesh level point clouds based on the first displacement vector, and determining the second displacement vector of the first metal mesh level point cloud based on the point-to-point matching result at the metal mesh level, includes: Based on the first displacement vector, the first metal mesh layer point cloud is deformed to obtain the initial metal mesh deformation point corresponding to each point in the first metal mesh layer point cloud. Based on the coordinates of the initial deformed metal mesh point cloud, a set of candidate neighborhood points for metal mesh layer point pair matching is determined in the second metal mesh layer point cloud for each initial deformed metal mesh point cloud. Based on the first metal mesh layer point cloud and the candidate neighborhood point set corresponding to each point in the first metal mesh layer point cloud for metal mesh layer point pair matching, the best matching point pair between the first metal mesh layer point cloud and the second metal mesh layer point cloud is obtained by nearest point filtering. The second displacement vector of the first metal mesh layer point cloud is determined based on the best matching point pair between the first metal mesh layer point cloud and the second metal mesh layer point cloud.
5. The method according to any one of claims 1-4, characterized in that, Point cloud registration is performed on the source point cloud data and the deformed point cloud data, including: The first anchor tray layer point cloud and the first metal mesh layer point cloud are used as non-redundant source point cloud data, and the second anchor tray layer point cloud and the second metal mesh layer point cloud are used as non-redundant deformed point cloud data. The RANSAC algorithm is used to perform coarse registration on the non-redundant source point cloud data and the non-redundant deformed point cloud data to obtain the initial estimated transformation matrix. The non-redundant deformed point cloud data is processed based on the initial estimated transformation matrix to obtain the initial estimated deformed point cloud; The Iterative Closest Point (ICP) algorithm is used to perform fine registration of the non-redundant source point cloud data and the initially estimated deformed point cloud to obtain a fine registration transformation matrix. The initial estimated deformed point cloud is processed based on the fine registration transformation matrix to obtain the registered deformed point cloud data.
6. A tunnel displacement field detection device, characterized in that, include: The acquisition module is used to acquire source point cloud data and deformed point cloud data of the target tunnel area, wherein the target tunnel area includes anchor trays and metal mesh; The hierarchical feature extraction module is used to construct the hierarchical features of the source point cloud data and the deformed point cloud data respectively, and obtain the first anchor tray hierarchical point cloud and the first metal mesh hierarchical point cloud in the source point cloud data, and the second anchor tray hierarchical point cloud and the second metal mesh hierarchical point cloud in the deformed point cloud data. An anchor tray level matching module is used to perform point cloud registration on the source point cloud data and the deformed point cloud data, and then perform point-to-point matching on the first anchor tray level point cloud and the second anchor tray level point cloud, and determine the first displacement vector of the first anchor tray level point cloud based on the point-to-point matching result of the anchor tray level. The metal mesh layer matching module is used to perform metal mesh layer point pair matching on the first metal mesh layer point cloud and the second metal mesh layer point cloud based on the first displacement vector, and to determine the second displacement vector of the first metal mesh layer point cloud based on the metal mesh layer point pair matching result. Other hierarchical matching modules are used to determine the third displacement vector of the point cloud other than the first anchor tray level point cloud and the first metal mesh level point cloud in the source point cloud data by using the first displacement vector and the second displacement vector through inverse distance weighted interpolation.
7. The apparatus according to claim 6, characterized in that, The hierarchical feature construction module is specifically used for: A KDTree spatial index is established, and a hybrid search strategy is used to determine the neighborhood point set of each point in the source point cloud data and the deformed point cloud data. The covariance matrix is calculated based on the neighborhood point set of each point in the source point cloud data and the deformed point cloud data. Spectral decomposition is performed on the covariance matrix to obtain the hierarchical features of each point in the source point cloud data and the deformed point cloud data; Based on the hierarchical features, the first anchor tray hierarchical point cloud and the first metal mesh hierarchical point cloud in the source point cloud data, as well as the second anchor tray hierarchical point cloud and the second metal mesh hierarchical point cloud in the deformed point cloud data, are determined.
8. The apparatus according to claim 6, characterized in that, The anchor tray hierarchical matching module is specifically used for: The first anchor tray level point cloud and the second anchor tray level point cloud are respectively aggregated to obtain the first anchor tray aggregation point corresponding to the first anchor tray level point cloud and the second anchor tray aggregation point corresponding to the second anchor tray level point cloud. The best matching point pair between the first anchor tray aggregation point and the second anchor tray aggregation point is obtained by the nearest point filtering; The first displacement vector of the first anchor tray level point cloud is determined by interpolation based on the displacement vector of the best matching point pair between the first anchor tray aggregation point and the second anchor tray aggregation point.
9. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1-5.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-5.